Correlation of pulmonary arsenic metabolism and toxicity

Item Type text; Dissertation-Reproduction (electronic)

Authors Barber, David Stewart, 1970-

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this materialis made possible by the University Libraries, University of Arizona.Further transmission, reproduction or presentation (such aspublic display or performance) of protected items is prohibitedexcept with permission of the author.

Download date 20/07/2021 08:46:37

Link to Item http://hdl.handle.net/10150/282551

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI

films the tact directly fi'om the original or copy submitted. Thus, some

thesis and dissertation copies are in typewriter face, while others may be

fi-om any type of computer printer.

The quality of this reproduction is dependent upon the quality of the

copy submitted. Broken or indistinct print, colored or poor quality

illustrations and photographs, print bleedthrough, substandard margins,

and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete

manuscript and there are missing pages, these will be noted. Also, if

unauthorized copyright material had to be removed, a note will indicate

the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by

sectioning the original, beginning at the upper left-hand comer and

continuing fi'om lefl to right in equal sections with small overlaps. Each

original is also photographed in one exposure and is included in reduced

form at the back of the book.

Photographs included in the original manuscript have been reproduced

xerographically in this copy. Higher quality 6" x 9" black and white

photographic prints are available for any photographs or illustrations

appearing in this copy for an additional charge. Contact UMI directly to

order.

UMI A Bell & Howell Information Company

300 North Zed) Road, Ann Aibor MI 48106-1346 USA 313/761-4700 800/521-0600

t X.

i I ?,

I

CORRELATION OF PULMONARY ARSENIC METABOLISM

AND TOXICITY

by

David Stewart Barber

A Dissertation Submitted to the Faculty of the

COMMITTEE ON PHARMACOLOGY AND TOXICOLOGY (GRADUATE)

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

19 9 7

UMX Number: 9814453

UMI Microform 9814453 Copyright 1998, by UMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103

THE UNIVERSITY OF ARIZONA ® GRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have

read the dissertation prepared by David S t e w a r t B a r b e r

entitled C o r r e l a t i o n o f P u l m o n a r y A r s e n i c M e t a h o T i g m

a n d T o x i c i t y

and recommend that it be accepted as fulfilling the dissertation

re q u i r e m e n t f o r t h e D e g r e e o f D o c t o r o f P h i l o s o n h v

9/9/97

?/f/f7 Dajie / ^

' R i c h a r a V a l l a i n c o u r t Date

M..CxiAisui 1/9/97 D e a n E . C a r t e r Date

Date

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

t. CjiJfiiJiy 9/fA Dissertation Director Date

D e a n F . C a r t e r

3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under the rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

ACKNOWLEDGMENTS

I could not have accomplished the work in this

dissertation without help from quite a few folks. First, I

would like thank my advisor. Dr. Dean Carter, my committee,

Dr. Tom McClure, and the members of the Carter Lab. These

people were instrumental in the day to day function of this

research. Next, I thank Mike and Rick (hope you get the big

one that got away), Ray and Leonard (hope you shoot that

elusive game in the eighties), and the members of "Bad

Habit" for keeping me sane. Finally, the greatest thanks go

to my wonderful wife, Carol, and our families who have been

loving, understanding, and supportive. This dissertation is

dedicated to them.

5

TABLE OF CONTENTS

Page LIST OF FIGURES 7

LIST OF TABLES 9

ABSTRACT 10

1. INTRODUCTION 11

1.1 Arsenic chemistry 11 1.2 Arsenic toxicology 13 1.3 Sources of airborne arsenic exposure 15 1.4 Arsenic metabolism 17 1.5 Lung structure and function 23 1.6 Absorption and disposition of

airborne arsenic 25 1.7 Metabolism of arsenicals by the lung 26 1.8 Pulmonary toxicity of inhaled arsenic....26 1.9 Arsenic and lung cancer 27 1.10 Types of pulmonary tumors caused by

arsenic 31

STATEMENT OF THE PROBLEM 32

HYPOTHESIS 33

RESEARCH OBJECTIVES 34

2. MATERIALS AMD METHODS 35

2.1 Metabolism experiments 35 Materials 35 Arsine generation 36 Redox incubations 36 Methylation incubations 39 Arsenite-glutathione complexation 42

2.2 Toxicity experiments 44 Cell and slice viability 44 Hsp32 induction 48 DNA single strand break assay 50

2.3 Modeling 52

3. RESULTS 57

3.1 Metabolism results 57

3.2 Toxicity results 91

3.3 Modeling results 102

DISCUSSION 109

4.1 Metabolism studies 109

4.2 Toxicity studies 127

4.3 Modeling and correlation of metabolism

toxicity 133

SUMMARY AND CONCLUSIONS

APPENDIX A

REFERENCES

138

140

145

7

LIST OF FIGURES

Fig\are Page

1. Metabolic scheme of arsenic in mammals 23

2. Inhibition of arsenite methylation in lung cytosol by PAD and SAH 58

3. Reduction of lOO^M As(V) by GSH, lung homogenates, and lung homogenate + GSH 60

4. Concentration dependence of As(V) reduction... 61

5. Oxidation of lOOuM As (III) 63

6. Concentration dependence of As(III) oxidation. 64

7. Loss of arsine from PBS and rat lung homogenates 66

8. Formation of As(III) from arsine in PBS and rat lung homogenates 67

9. Formation of As(V) from arsine in PBS and rat lung homogenates 68

10. Loss of arsine from PBS and guinea pig lung homogenates 70

11. Formation of As (III) from arsine in PBS and guinea pig lung homogenates 71

12. Methylation of As(IlI) in rat lung cytosol.... 73

13. Lineweaver-Burke plot of As(III) methylation data 74

14. pH dependence of As(III) methylation in rat lung cytosol 75

15. Methylation of As(V) in rat lung cytosol 76

16. Methylation of arsine in rat lung cytosol 77

17. pH dependence of arsine methylation in rat lung cytosol 78

18. Speciation of methylated derivatives of arsine 79

8

Figure Page

19. Methylation of MMA by rat lung cytosol 80

20. Non-protein sulfhydryls in rat lung homogenates treated with arsenicals 83

21. Mass spectrum of As(SG)3 standard 85

22. Daughter ion spectrvim of 994 ion from AS(SG)3 standard 86

23. Formation of As(SG)3 in rat lung homogenates.. 87

24. Time course of As(SG)3 formation from As(III) and As (V) 89

25. Methylation of As(SG)3 by rat lung cytosol.... 90

26. Toxicity of arsenicals in BEAS-2B cells treated for 24 hours 92

27. Time course of LDH release from BEAS-2B cells treated with arsenicals 93

t,

^ 28. Toxicity of arsenicals in hamster lung slices. 94 I I 29. Arsenic accumulation in hamster lung slices i treated with arsenicals for 24 hours 96

30. Histology of hamster lung slices treated with arsenicals for 24 hours 97

31. Hsp32 induction in BEAS-2B cells treated with arsenicals for 4 hours 99

3 r

\ 32. Concentration dependence of hsp32 induction I by arsine and arsenite 100

I 33. Induction of DNA single strand breaks by ! arsenicals 102

34. Modeling of As(V) metabolism 103

35. Modeling of As (III) metabolism 104

36. Modeling of ASH3 metabolism 105

37. Effect of pH on reduction potential of As(V)..117

9

LIST OF TABLES

Table Page 1. Percentage 1, 10, and 100|iM As(V) reduced

by GSH and rat lung homogenates 62

2. Percentage 1, 10, and 100|aM As (III) oxidized in PBS and rat lung homogenate 65

3. NPSH content of aqueous solutions of As(SG)3 (showing dissociation of complex under assay conditions) 82

4. Predicted arsenic metabolite concentrations in toxicity assays 107

10

ABSTRACT

In lung preparations, As(V) was reduced to As(III)

[first order rate constant of 0.0104/min]; As(III) was

oxidized to As(V) [first order rate constant of 0.005/min],

methylated to MMA [Kin=5.383nM, Vmav= 0.00031

Hmol/liter/min/mg], and complexed with GSH; MMA was

converted to DMA [Km=63.4 |lM, 0.0000384

(imol/liter/min/mg] ; and arsine was oxidized to As (III) and

As(V) and methylated.

Toxicity of As(III), As(V), MMA, DMA, and arsine was

assessed by measuring effects on cell and slice viability,

hsp32 induction, and production of DNA single strand breaks.

Because all species of arsenic did not produce effects, it

was possible to deduce an "active" form of arsenic from

these studies.

Pulmonary arsenic metabolism was modeled using

SIMUSOLV. This model indicated that arsine disposition

cannot be explained solely by oxidation to As(III) before

methylation or further oxidation occurs. The concentration

of arsenic species present in toxicity studies were

predicted with this model and correlated to observed

effects. There was good correlation between reduction of

As(V) to As (III) with toxicity and hsp32 induction.

However, the effects observed for arsine did not correlate

with oxidation to arsenite.

11

CHAPTER 1

INTRODUCTION

1.1 Arsenic chemistry. Arsenic, atomic number 33, is

a group V metalloid with a molecular weight of 74.9216 g/mol

(Weast, 1976). It exists in formal oxidation states of

(-III), 0, (+III), and (+V). Arsenic forms inorganic and

organic compounds. Inorganic forms include the arsenites,

arsenates, and arsenides. Organic forms include

monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA),

which are the major arsenic metabolites in humans, as well

as arsenobetaine, arsenocholine, and trimethylarsine oxide,

which are foirmed by shellfish, fungi, and some rodents (Oya-

Ohta et al., 1996; Lau et al., 1987)

Inorganic pentavalent arsenic (arsenic acid, H3ASO4) is

the most oxidized form of arsenic. It is thermodynamically

stable and is the most prevalent form of arsenic in the

environment. Salts of this compound are referred to as

arsenates. Arsenates have pKa values of 2.20, 6.97, and

11.53, thus are ionized at physiological pH. This has

significance in transport of arsenic, as arsenates cannot

diffuse into the cell, but must gain access via a

transporter. Arsenate is similar to phosphate in chemical

structure and charge. Arsenate mimics phosphate in

biological reactions, but forms unstable bonds that

spontaneously hydrolyze.

12

Inorganic trivalent arsenic (arsenous acid, HASO2)

forms salts called arsenites. Trivalent arsenicals are

hydrated in solution to form As (OH) 3. With pKas of 9.1,

12.13, and 13.40, uncomplexed arsenite is an uncharged

molecule at physiological pH. This allows diffusion of

arsenite across biological membranes. Arsenite has a high

affinity for reduced thiol groups and produces toxicity by

binding to thiols.

Metallic, or elemental, arsenic is found naturally in

the ores realgar and orpiment. Elemental arsenic decomposes

at 613°C to form AS2O3 (Weast, 1976). This occurs during

ore smelting and is used to prepare arsenic for other uses.

The extent of exposure and toxicity of elemental arsenic are

unknown.

Arsenic in the (-III) oxidation state forms arsines and

arsenides. Arsenides, such as gallium arsenide, are usually

solids and are used in semi-conductors, lasers, and solar

cells. Arsine (ASH3) is the gaseous hydride of arsenic. It

is a colorless, non-irritating gas (B.P. -62.5®C) with a

reportedly garlic-like odor (Hocken and Bradshaw, 1970). It

is soluble in organic solvents and slightly soluble in

aqueous solution (8.93mM, Weast, 1976). Arsine is formed

from the combination of arsenic, acid or base, and an

elemental metal in the following reaction:

6H'^ + 2A1(0) + HASO2 > AsH^ + 2H2O + 2Al(III)

13

This provides ample opportunity for accidental exposure

to arsine in industrial settings where these ingredients are

common. Arsine is used as a dopant for silicon based chips

and growing gallium arsenide crystals in the electronics

industry.

Toxicologically relevant methylated arsenicals contain

pentavalent arsenate and are acids in solution.

Monomethylarsonic acid (MMA, CH3AsO(OH)2) has pKas of 4.1

and 8.7. It is a charged species at physiological pH.

Dimethylarsinic acid (DMA, (CH3) 2As (O) OH) has a pKa of 6.2.

DMA is also charged at physiological pH.

1.2 Arsenic toxicology. Arsenic has a long

toxicological history. It was used as a poison and medicine

by the Greeks and Romans 2400 years ago (Gilman et al.,

1985), exemplifying the famous statement by Paracelsus: "All

things are poison..., solely the dose determines that the

thing is not a poison." Exposure to high doses of arsenic

(100+ mg) can be fatal. Acute arsenic exposure causes

fever, hepatomegaly, melanosis, cardiac arrhythmia, upper-

respiratory tract problems, peripheral neuropathy,

gastrointestinal distress, mucous membrane damage, and

hematopoietic effects (anemia and leukopenia) (Klaassen,

1996) . High doses of arsenic can also cause fetal

malformations, especially damage to the neural crest cells

(Shalat et al., 1996).

14

Chronic exposure to arsenic causes neurotoxicity, liver

injury (jaundice, cirrhosis), peripheral vascular disease

(blackfoot's disease), and cancer. Arsenic is classified as

a hvunan carcinogen by the EPA and lARC based on sufficient

human epidemiological evidence. Inhalation causes lung

cancer. Ingestion causes skin cancer, hemangiosarcoma of

the liver, lymphoma, leukemia, kidney, and bladder cancer.

1.2.1 Mechanisms of arsenic toxicity. Arsenic toxicity

is dependent on the form of arsenic. Arsenates are

uncouplers of oxidative phosphorylation. They produce

toxicity by substituting for phosphate in biochemical

reactions (Squibb and Fowler, 1983). The arsenate-phosphate

bond is unstable and hydrolyzes rapidly, in a process termed

arsenolysis (Aposhian, 1991). When this occurs during ATP

synthesis, ADP-arsenate is formed (Gresser, 1981). ADP-

arsenate undergoes arsenolysis and depletes the high energy

phosphate bonds necessary for cell viability.

Arsenite has a high affinity for thiols. Many enzymes

contain thiols in their active sites. By binding to these

thiols arsenite inhibits cellular functions, including

respiration and metabolism (Squibb and Fowler, 1983; Webb,

1966). The classic example is the inhibition of pyruvate

dehydrogenase (PDH) by arsenite reported by Stevenson et al.

(1978). The lipoic acid subunit of PDH contains vicinal

15

thiols that bind arsenite to form a stable six-membered ring

and inhibit the enzyme.

Methylated arsenicals have very low acute toxicity when

compared to inorganic arsenicals due to their low affinity

for macromolecules. For this reason, formation of MMA and

DMA from inorganic arsenic is considered a detoxification

step(Klaassen, 1990). It is important to note that As(III)

forms of these compounds can form and these will react

differently than the As(V) fomns (Cullen et al., 1984).

Arsine is the most acutely toxic form of arsenic. The

threshold limit value (TLV) for arsine is O.OSppm for an 8-

hour workday (ACGIH, 1982). The major symptom of arsine

poisoning is massive hemolysis (Pernis and Magistretti,

1960; Fowler and Weissberg, 1974). Renal (Meuhrcke and

Pirani, 1968), cardiac (Josephson et al., 1951), hepatic

dysfunction, peripheral nervous system damage and pulmonary

edema (Hocken and Bradshaw, 1970) are also observed in cases

of arsine exposure. The mechanism of arsine toxicity is

unknown, but has been hypothesized to result from oxidative

damage (Labes, 1937); depletion of reduced glutathione

(Blair et al., 1990b; Pernis and Magistretti, 1960); and

inhibition of Na+/K+ ATPase (Levinski et al., 1970).

1.3 Sources of airborne arsenic. This dissertation

deals with the effects of arsenic on the lung, so exposure

by inhalation is the only exposure route that will be

16

considered. Arsenic has been utilized for glassmaking,

herbicides, pesticides, wood preservatives, pigments and

semi-conductors. It is also liberated during smelting and

fossil fuel combustion. Most airborne arsenic is in the

inorganic form. Arsenic is carried on the surface of

particles composed of other materials. Airborne arsenic

concentrations from O.lng/m^ in uninhabited areas

(Antarctica) to 500ng/m^ around copper smelters have been

reported by Rabano et al. (1989). The average arsenic

content of US urban air is 20ng/m^ (Rabano et al., 1989;

Davidson et al., 1985). This translates to an average daily

exposure of 400ng arsenic/day by inhalation in most cities,

assuming 100% deposition.

The form of arsenic released after combustion is

unclear but has been hypothesized to be AS4O6 (a gas phase

form of arsenic(III) oxide). Reaction of liberated

arsenicals with air should lead to rapid oxidation and a

preponderance of As(V). There have been several studies

that speciated arsenic in ambient air. Rabano et al. (1989)

collected air samples in Los Angeles, California and

separated the samples into <2.5/im and >2.5/im particles.

Speciation of the arsenic present in each particle size

revealed that the ratio of As (III)/As (V) was about 1.5 in

small particles and 0.9 in larger particles. These results

differ from those found by Solomon (1984) in Tucson,

Arizona. The As (III)/As(V) ratio in that study was 0.3.

17

Arsine is a gas, so exposure is almost exclusively by

inhalation. Arsine forms readily from mixtures of arsenic,

acid or base, and elemental metals (see reaction on p. 12).

Though these conditions are commonly found in industrial

settings, there have been no measurements of arsine in

workplace air.

1.4 Arsenic metabolism. The metabolism of arsenic

occurs via reduction, oxidation, and methylation. Arsenic

undergoes two electron reduction or oxidation allowing

interchange between trivalent and pentavalent forms.

Reduction and oxidation are highly dependent on pH and

oxygen content. Arsenate is the thermodynamically stable

form of arsenic and is prevalent except at low pH and/or low

oxygen environments. At physiological pH, in systems

containing oxygen, oxidation of arsenous acid, (As(III), to

arsenic acid, As(V), should occur spontaneously as predicted

by the following equation:

HASO2 + 2H2O > H3ASO4 + 2H* + 2e"

E0=-0.56v (Dean, 1979)

Reduction of As(V) will occur in the presence of

reduced thiols. This reaction is depicted in the following

equation:

H3ASO4 + 2RSH > HASO2 + RSSR + 20H"

18

This reaction has been observed in aqueous solutions

(Scott et al, 1993; Delnomdedieu et al, 1994). Cullen et

al. (1984) observed the same reaction with methylated

arsenicals (MMA and DMA). He reported reduction potentials,

e, of -0.229V and -0.233V for cysteine and glutathione,

respectively.

The other toxicologically relevant form of arsenic is

arsine. The arsenic in arsine is thought to be As (-III).

Since this is the fully reduced form of arsenic it will be

oxidized in the presence of oxygen. Arsine may oxidize in

several ways. One is to produce superoxide and arsenous

acid (As(III)) as illustrated in this equation:

ASH3 + 6O2 + 2H2O > 6O2" + + HASO2

The standard reduction potential,Eq', for this reaction

at physiological pH was calculated to be +0.31V by Rosner

(1989). Another is to react with water as depicted in this

reaction:

ASH3 + 2H2O > HASO2 + + 6e"

Eo=+0.189V (Dean, 1979).

In biological systems, arsine will react with cellular

macromolecules, possibly oxidizing in a series of two

electron steps, As(-III) to As(-I) to As(+I) to As(+III).

19

As arsine oxidizes, something must be reduced. In cells,

the electron acceptor could be oxygen, water, or biological

molecules. Strangely enough, addition of arsine to

hemoglobin solutions produces oxidation of heme, not

reduction (Hatlelid et al., 1995)

In vivo reduction of arsenic has been clearly

demonstrated by the presence of trivalent arsenic in the

blood and urine of animals dosed with pentavalent arsenic

(Ginsberg, 1965; Lerman et al, 1983; Vahter and Envall,

1983; Rowland and Davies, 1982). Rowland and Davies

(1982) showed that reduction occurred rapidly, as As(III)

was present in blood 5 minutes after an intraintestinal dose

of As(V), though this may be partially attributed to gut

microflora. The mechanism of reduction is unclear from in

vivo studies. In vitro reduction by human erythrocytes

indicates that reduction occurs by a thiol and protein

dependent reaction that may be enzymatic (Winski and Carter,

1995). Further reduction of arsenite to arsine requires a

strong reductant and is not likely to occur in biological

systems.

In vivo oxidation is demonstrated by experiments in

which arsenite was administered and the presence of arsenate

was observed at later timepoints (Bencko et al., 1976;

Lindgren et al., 1982; Vahter and Envall, 1983; Rowland and

Davies, 1982). Similar work showed that humans also oxidize

20

arsenic (Mealey et a.1., 1959). Oxidation of arsenite would

be expected to occur in the lung due to high oxygen tension.

The methylated metabolites of arsenic,

monomethylarsonic acid and dimethylarsinic acid, are organic

forms of pentavalent arsenic. Arsenic is methylated by the

enzymatic transfer of a methyl group from S-adenosyl-L-

methionine (SAMe) to arsenite. According to the methylation

reaction described by Challenger (1945), arsenic loses a

proton to become negatively charged and reacts with a

positively charged methyl group from SAMe, as depicted in

the following reaction:

•""CHa

As (OH) 3—> + {HO)2AsO~ > CH3ASH2O3

This is an oxidative methylation, so the arsenic must be in

the +(III) state to be methylated.

It is possible that the methyl group could be

negatively charged (~CH3). If arsine were to lose a hydride

ion to become positively charged, it could react with a

negatively charged methyl group. This could lead to non-

enzymatic methylation of arsine via the following reaction:

"CH3

ASH3 > •*"ASH2 + H~ > CH3ASH2

21

Similar nonenzymatic methylation of mercury has been

observed in the presence of methylcobalamin (Choi et al.

(1994).

Very little work has been done on the methylation of

arsenic by the lung. Georis et al. (1990) reported that

lung slices produced only 30% of the methylated arsenic

produced by the liver. Organic forms of arsenic have low

affinity for macromolecules and are readily excreted.

Therefore, methylation is generally considered a

detoxification pathway in arsenic metabolism. If the lung

is actually deficient in arsenic methylation, it could be

predisposed to toxicity from arsenic.

In vivo, metals exists as complexes with physiological

ligands and not as free ions. Arsenic is no different,

although little attention has been paid to the actual

intracellular species of arsenic. Glutathione (GSH, y-

glutamylcysteinylglycine), a tripeptide found at millimolar

concentrations in most cells of the body is an attractive

ligand because it has an available thiol group that can

react with arsenicals. As early as 1924, Voegtlin et al.

recognized that GSH reacted with arsenic. Glutathione and

arsenic interact in several ways: reduction of arsenate to

arsenite with concomitant formation of GSSG and formation of

arsenic-glutathione complexes.

22

As(V) + 2 GSH > As(III) + GSSG

As (III) + 3 GSH > As(SG)3

These reactions have been observed in aqueous solution

(Delnomdedieu et al, 1994; Delnomdedieu et al, 1993; Cullen

et al, 1984; Scott et al, 1993). Scott et al. (1993)

demonstrated that various arsenicals could form complexes

with glutathione, including arsenite, MMA, and DMA. This is

not surprising given the affinity of As(III) for thiols.

These complexes are very labile, making detection very

difficult. NMR studies by Delnomdedieu et al. (1994) showed

that the complex is stabilized under acidic conditions, but

dissociates at higher pH, releasing reduced GSH.

Recently, these reactions have also been found to have

biological significance. Styblo and Thomas (1995) found

that addition of preformed As(SG)3 inhibited glutathione

reductase. Delnomdedieu et al. (1994b) used NMR to show the

formation of As(SG)3 complex in rabbit erythrocytes treated

with arsenite. However, no one has isolated and quantified

the AS(SG)3 complex from a biological system.

23

OH OH

HO A* OH HO At(V) CH3

11 H O — — C H 3

11 O

Ancute MoaofnethyUnonic Acid

O O Dimethylaninic Acid

"y H

HO A« — OH

Aiicnitc

HO AjCDI) CH3

Mononiethylinenic(III) [MMAi(in)]

HO AJOII) — CH3

DinicthyIanenic(III) [DMAsam]

SG A

GS Ai SG

Aijcnitc^lutathione

H—Aj — H

Aninc

Complex [As(SG) 3]

Fig. 1 Metabolic scheme for arsenic in most mammals. As(SG)3, MMAs(III), and DMAs (III) have been identified in aqueous systems but have never been isolated from biological samples.

1.5 Lung Structure and Function. The lung is in direct

contact with ambient air. This allows the lung to exchange

gases, maintaining the oxygen supply to other tissues. It

also exposes the lung to a great deal of xenobiotics, both

gaseous and particulate. The large surface area of the lung

required for efficient gas exchange also makes the lung very

efficient at absorbing xenobiotics.

However, the lung is not a big bag. It is composed of

more than 40 cell types. These cells are arranged to form

the trachea and bronchi that are the large airways. These

airways bifurcate approximately 30 times, becoming smaller

and smaller until they end up in the acinar region (alveoli)

2 4

of the distal lung. The major epithelial cell types in the

upper airways are the ciliated columnar cells, non-ciliated

"brush" cells and the goblet and serous (mucous

secreting)cells. The major cell types in the distal acinar

region of the lung are the type I and II epithelial cells

and alveolar macrophages. Clara cells (non-ciliated, low

columnar cells) are found in the bronchioles and respiratory

bronchioles. The distribution of these cells changes as one

moves distally through the lung reflecting the changing

function of the lung.

In the large upper airways, no gas exchange occurs and

the major function is to conduct air to the distal lung and

to remove inhaled debris. Goblet and serous cells produce

mucous which traps particles and is constantly moved out of

the lungs by the action of the ciliated cells. As one moves

distally in the lung, cell distribution changes to reflect

function. The major function of the distal lung is gas

exchange and surfactant production. Type I epithelial cells

are the major cells involved in gas exchange, while Clara

and type II cells secrete surfactant proteins. The

phospholipid portion of surfactant is secreted primarily by

Type II cells. Alveolar macrophages are also much more

prevalent in the distal lung and scavenge small particles

and bacteria that get into the acinar region. Clara cells

are the most metabolically active cells of the lung as they

2 5

contain the highest P-450 content. The complexity of the

lung provides xenobiotics a number of different targets.

1.6 Absorption and disposition of airborne arsenic by

the lung. Disposition and absorption of inhaled arsenic

will depend largely on particle diameter. 23% of particles

in arsenic polluted air were larger than 5.5nm (Pinto and

McGill, 1953). These particles will impact on the

nasopharyngeal region and be cleared rapidly by mucociliary

action. Much of this arsenic will be swallowed and may be

absorbed in the gastrointestinal tract but is not of

importance in inhaled arsenic exposure. In 1974, Davison et

al. reported that much of the arsenic in coal fly ash was

found in particles of l-2|i.m diameter. These particles are

small enough to be carried deep into the lung before they

deposit.

The absorption of arsenic from deposited particles is

not well quantified. Absorption depends largely on the

solubility and size of the particles. Webb et al. (1987)

showed that decreasing the mean particle volume of GaAs

particles greatly increased the absorption of arsenic from

the lungs. Inamasu et al. (1982) found that calcium

arsenate (slightly soluble arsenical) was retained much

longer in rat lungs after instillation than arsenic trioxide

(soluble arsenical).

2 6

1.7 Metabolism of arsenicals by the lung. Little work

has been done on the ability of the lung to metabolize

arsenic. Work by Georis et al. (1990) showed that lung

slices from hamsters methylated arsenic very slowly compared

to liver. This makes it tempting to hypothesize that the

lung is at risk due to its inability to detoxify arsenic by

methylation. There have been no studies which measured the

reduction or oxidation of arsenicals by the lung. There

have also been no studies which determined the metabolism of

arsine or arsenic complexation with glutathione by the lung.

The ability of the lung to metabolize arsenic must be

quantified in order to understand the effects produced by

various arsenicals.

1.8 Pulmonary toxicity of inhaled arsenic. Inhalation

of arsenicals produces effects on the lung other than

cancer. Occupational exposure to arsenic trioxide dusts

causes nasal irritation (Morton and Caron, 1989) and high

doses can cause septal perforation (Pinto and McGill, 1953),

but does not appear to cause respiratory impairment (Perry

et al., 1948). Intratracheal instillation of 13 mg As/kg

arsenic trioxide caused irritation and hyperplasia in the

lungs of rats (Goering et al., 1988). Inhalation of very

high concentrations of MMA and DMA (<2000 mg As/m3) caused

respiratory distress and death in mice and rats (Stevens et

al., 1979).

2 7

Arsenic also produces effects on the immune system.

Inhalation of arsenic trioxide caused increased

susceptibility to respiratory infections as a result of

injury to alveolar macrophages (Aranyi et al., 1985).

Instillation of as little as 1 mg As/kg in rats suppressed

tumor necrosis factor (TNF) production by pulmonary alveolar

macrophages (PAM) 24 hours after exposure (Lantz et al,

1994) .

1.9 Arsenic and Lung Cancer. Arsenic was suspected as

a carcinogen as early as 1820 (Paris, 1820). In 1879,

inhaled arsenic was suggested to cause lung cancer in

miners, but not until the 1930's did causal evidence arise

(Montgomery, 1935; Neubauer, 1947). There is a clear link

between inhaled arsenic and lung cancer. Numerous studies

have found increased incidences of lung cancer in employees

of smelters (Lee and Fraumeni, 1969; Axelson et al., 1978;

Welsh et al., 1982; Lee-Feldstein, 1986; Enterline et al.,

1987; Jarup et al., 1989) and pesticide production (Ott et

al., 1974). These studies found a dose-response

relationship between arsenic exposure and lung cancer. The

relative risk of developing lung cancer was 3 (Lee and

Fraumeni, 1969) to 8.7 times(Jarup and Pershagen, 1991)

greater in arsenic exposed individuals than in controls.

These studies also found that there were long latency times

2 8

in the induction of cancer, ranging from 34-51 years (Lee

and Fraumeni, 1969; Axelson et al., 1978).

Attempts to model arsenical induced respiratory cancer

in animals have been largely unsuccessful. Studies by

Ishinishi et al. (1983) and Pershagen et al. (1985) indicate

that intratracheal instillation of arsenicals alone can

produce cancer in animals. In other studies in which

instilled arsenic produced cancer, arsenic was co­

administered with other compounds, such as benzo(a)pyrene

(Pershagen et al., 1984a) and charcoal/sulfuric acid.

The form of arsenic that causes cancer is unknown.

Because most industrial arsenic exposures are thought to be

AS2O3, arsenite is proposed to be the active form. Most

studies utilize some form of trivalent arsenic and attribute

effects to As(III) . However, metabolism of arsenic occurs

in most mammals and it is possible that observed effects are

due to other arsenic species.

There have been no studies on the carcinogenicity of

organic arsenicals. However, Yamanaka et al. (1989, 1993,

1995) have shown that DMA is capable of producing DNA

lesions (strand breaks, crosslinks) specifically in the lung

after in vivo treatment.

Arsenic clearly causes cancer, however debate rages

over the mechanism of carcinogenicity. Cancer appears to

develop by a series of steps. The multi-stage model of

carcinogenesis originally described by Berenblum and Shubik

2 9

(1947), contained two steps: initiation and promotion.

This was expanded to contain a third stage, progression,

after work by Boutwell (1964) and Foulds (1965). There are

specific criteria for a compound to act at each of these

stages. Pitot and Dragan (1994) described the events that

occur during each stage of carcinogenesis. Initiation

involves damage to DNA that results in mutations which are

fixed in the genome. Initiation is irreversible, except by

cell death and has no dose-response threshold. Mutations in

proto-oncogenes and genes responsible for signal

transduction are especially important. Promotion involves a

selective proliferation of initiated cells. Promotion is

reversible and requires the continued administration of the

promoter for activity. Promotion also exhibits a threshold

dose-response effect. Promoters are often agents that cause

proliferation but do not damage DNA. Promoters may act by

altering gene expression or inhibiting apoptosis.

Progression is the alteration of tumor cells from benign to

malignant, with associated increases in growth rate,

invasiveness and altered morphology. Progression is

irreversible and involves complex genetic changes such as

chromosomal aberrations and gene amplification. This

results from karyotypic instability and clastogenesis.

Where does arsenic fit into this model? Arsenic

produces little or no response in mutagenicity tests

(Jacobson-Kram and Montalbano, 1985). Arsenite can act as

3 0

a co-mutageri/ increasing mutations caused by other compounds

such as ultraviolet light (Li and Rossman, 1989) and MNNG

(Nunoshiba and Nishioka, 1987). It has been hypothesized

that arsenite acts as a co-mutagen by inhibiting DNA repair

(Rossman, 1981). Recently, arsenite has been shown to

specifically inhibit DNA ligase II (Li and Rossman, 1989).

In light of these studies, inorganic arsenic is unlikely to

be an initiator, but may increase the efficacy of other

initiators, acting as a co-initiator. It is also possible

that metabolism to DMA, which Yamanaka et al. (1989) has

shown to produce DNA damage at high concentrations, could

cause initiation.

Several recent studies claim that DMA acts as a tumor

promoter in kidney (Wanibuchi et al., 1995) and lung

(Yamanaka et al., 1996). As(III), As(V), and

dimethylarsenic have been shown to act as tumor promoters in

rats treated with diethylnitrosamine (Shirachi et al. 1987).

Sodium arsenite has been shown to increase mitogen

stimulation in fibroblasts, leading to increased

proliferation (Van Wijk et al., 1993). Cell proliferation

is a hallmark of tumor promoters. There is evidence that

both inorganic and organic arsenic may act at the promotion

stage of carcinogenesis.

Inorganic arsenicals cause chromosomal aberrations and

sister chromatid exchange (Lee et al., 1985; Jacobson-Kram

and Montalbano, 1985). Arsenic causes micronuclei formation

3 1

in bladder (Moore et al., 1997) and mouse bone marrow cells

in vitro (Tinwell et al., 1991). The clastogenicity of

arsenic has been attributed to its inhibition of DNA ligase

(Li and Rossman, 1989) . Arsenicals also produce

amplification of the dihydrofolate reductase gene (Lee et

al., 1988) and over-expression of c-fos (Gubits, 1988).

Clastogenicity and gene amplification are hallmarks of tumor

progressors. These studies indicate that arsenicals may be

active in all three stages of carcinogenesis.

1.10 Types of pulmonazY tumors caused by arsenic.

Several studies have examined the types of tumors induced by

inhaling arsenic. The premise being that if there were a

predominance of certain types of tumors that may indicate a

susceptible type of cells. The results of these studies are

inconclusive. Newman et al. (1976) reported a predominance

of epidermoid carcinomas among cases of bronchogenic

carcinomas in smelter workers. Wicks et al. (1981) reported

a high percentage of adenocarcinomas. Pershagen et al.

(1987) found that arsenic caused no changes in the types of

tumors observed in smokers and produced tumor types similar

to smokers in non-smokers. This indicates that arsenic does

not target a specific type of cell in the lung for damage.

3 2

Statement of the Problem

Arsenic is emitted during fossil fuel combustion, ore

smelting, and semi-conductor manufacture. Exposure by

inhalation is widespread and is especially common in

industrial settings. Epidemiological studies clearly show

that inhalation of arsenic produces lung cancer; while

ingestion of arsenic leads to skin, liver, and bladder

cancer. The mechanism of arsenical carcinogenesis is not

understood. Unfortunately, arsenic toxicology in humans is

not modeled well by animal experiments. By any measure

humans are more susceptible to arsenic toxicity than

animals.

Arsenic toxicology is complicated by metabolism in most

mammals, so inhalation of one form of arsenic actually

results in exposure to multiple forms of arsenic. The

accepted metabolic pathway in humans is arsenate (+V)—>

arsenite (+III) —> monomethylarsonic acid (MMA) —>

dimethylarsinic acid (DMA). Inorganic arsenicals are more

acutely toxic than methylated derivatives. This has led to

the belief that methylation is a detoxification pathway.

Whether this is true for chronic exposures and the

development of cancer is unknown.

The lung is the site of exposure to airborne arsenic,

as well as the target organ, so other organs are probably

3 3

not involved in the pulmonary toxicity of inhaled arsenic.

This unusual situation makes understanding the distribution,

metabolism, and toxicity of arsenicals in the lung critical

to understanding the effects of arsenicals on the lung. In

spite of the need for information on these subjects, little

work has been done to determine the ability of the lung to

metabolize arsenic or the effects of different forms of

arsenic in the lung.

Because metabolism creates exposure to several forms of

arsenic simultaneously, any risk assessment which does not

identify the toxic form of arsenic is flawed. It is probable

that certain toxicities have been associated with the wrong

form of arsenic. As this could have a large impact on risk

assessment guidelines for inhaled arsenic, it is necessary

to develop a more complete understanding of the metabolism

and toxicity of arsenicals in the lung. This work addresses

this need by determining the metabolism of arsenic in the

lung and correlating observed effects with the concentration

of arsenic species produced by metabolism.

General Hypothesis

Arsenic is metabolized by the lung. Arsenic toxicity

in the lung is dependent on chemical form. Toxicity of

arsenicals can be correlated to the arsenic species produced

by metabolism.

3 4

Research objectives

1. Metabolism of arsenic by the lung

•Quantify reduction, oxidation, and methylation

capacity of the lung

•Isolate and quantify arsenic complexes with

glutathione

2. Toxicity of arsenic in the lung

•determine acute toxicity of arsenicals in in vitro

systems.

•determine effect of arsenicals on gene expression and

DNA damage.

3. Model metabolism of arsenic in the lung

•Correlate toxicity of arsenicals to metabolism

•Detemnine if effects produced by arsine can be

explained solely by oxidation to arsenite.

3 5

CHAPTER 2

MATERIALS AND METHODS

2.1.1. Materials for metabolism experiments.

Sodium arsenite was purchased from Fisher Chemical Co.

(Fairlawn, NJ, S-225, Lot# 713607). Arsenite was

dissolved in PBS and adjusted to pH 7 with IN HCl.

Sodium arsenate (heptahydrate) was purchased from J.T.

Bcdcer (Phillipsburg, NJ, 1-3486, Lot# 428273). Arsenate

was dissolved in PBS and the pH adjusted to 7.0 if

necessary. Monomethylarsonic acid (MMA, Disodium) was

purchased from Pfaltz & Bauer, Inc. (Stamford, Conn,

S06090). Dimethylarsinic acid (DMA, cacodylic acid,

sodium salt) was purchased from Sigma Chemical Co. (St.

Louis, MO., C-0250) . ^^As(V) was purchased from Los

Alamos National Laboratories as a solution in IN HCl.

This stock solution was diluted to ImL with pH 7.4

phosphate buffer upon receipt to inhibit auto reduction.

S-Adenosyl-L-[methyl-^H] Methionine (^H-SAMe) was

purchased from Amersham Life Science (#TRK581,Arlington

Heights, IL) .

Ketamine-Xylazine-Acepromazine (KRA) is a mixture of

40 mg/mL Ketamine (Ketaset®, Fort Dodge Labs, Inc., Ft.

Dodge, lA.), 5 mg/mL Xylazine (Rompun®, Miles, Inc.,

Shawnee Mission, KS.), and 2.5 mg/mL Acepromazine

(Acepromazine Maleate, Fermenta Animal Health Co., Kansas

3 6

City, MO.)' Animals were anesthetized using

intramuscular injections of ImL KRA/kg body weight.

2.1.2. Arsine generation. Arsine was made by the

reaction of sulfuric acid with zinc arsenide. An aqueous

slurry of zinc arsenide (approx. 0.5g in 2-3mL H2O

[Pfaltz and Bauer, Inc., Waterbury, CT.) was reacted with

50% sulfuric acid (added in 1 mL increments until desired

arsine concentration was reached). Evolved arsine was

bubbled into PBS using nitrogen as a carrier gas.

Aqueous concentration of arsine was determined by mixing

150 mL dosing solution with 1.35 mL 0.55%

silverdiethyldithiocarbamate (Eastman Kodak Co.,

Rochester, NY, #7464) in pyridine. Arsine was quantified

by absorbance at 510nm on a DU-7 spectrophotometer

(Beckman Instrument Co., ).

2.1.3. ^^As(V) reduction. For oxidation studies,

^^As(V) was reduced to ^^As(III) as described by Reay and

Asher (1977). Reducing solution consisted of 1 mL Milli-

Q water, 0.0187 gm sodium metabisulfite, 133.3 mL 1%

sodium thiosulfate (in water), and 14.2 mL concentrated

sulfuric acid. Reducing solution was mixed 1:2 with

arsenate solutions (10 nL ^^As(V) was mixed with 20 nL

arsenate) and incubated in the dark for 2 hours. After

3 7

incubation, pH of reduced solution was adjusted to 7.0

with NaOH. Neutralized solution was loaded on Bond Elut

strong anion exchange (SAX) columns (Varian, Harbor City,

CA, 1210-2017) pre-wet with 1 mL of methanol and

equilibrated with water. As(III) is not charged at

neutral pH and was eluted with 2 mL of H2O.

2.1.4. Lung homogenate preparation. Male Sprague-

Dawley rats (250-350g) were anesthetized by KRA injection

and killed by exsanguination (cutting the inferior vena

cava). Lungs were perfused through the left ventricle of

the heart with cold saline solution (40mls) to remove

blood from the lungs. Lungs were removed intact,

weighed, and diced. Lungs were then homogenized in 4x

weight volumes of PBS using 6 passes with a teflon glass

homogenizer. This resulted in a 20% (w/v) homogenate.

2.1.5. Inhibition of methylation with PAD/SAH.

Reduction and oxidation studies were carried out in rat

lung homogenates in which methylation was inhibited by a

mixture of periodate oxidized adenosine (PAD) and S-

adenosyl-homocysteine (SAH). PAD is a general

methyltransferase inhibitor that works by inhibiting the

SAH hydrolase and causing SAH to build up in the

incubation. Increased concentrations of SAH inhibit many

methyItransferases. PAD was prepared from adenosine by

3 8

the method of Khym and Cohn (1960) . SAH was purchased

from Sigma Chem. Co. A combination of 100 |iM PAD plus 1

mM SAH was used for all reduction/oxidation assays to

prevent methylation of arsenic. 0.25 mL of homogenate

was mixed with 0.25 mL of 2x arsenic and incubated at

ST'C.

2.1.6. Inorganic arsenic speciation. With

methylation inhibited, only As(III) and As(V) were

present. These arsenic species can be separated using

anion exchange chromatography as described by Winski and

Carter (1995). To prepare these columns, anion exchange

resin (Bio-Rad AG 1X8 chloride form, 100-200 mesh, #140-

1441, Bio-Rad Labs, Hercules, CA) was washed in 0.5N HCl

and then rinsed with Milli-Q water until the pH was

between 4-6. This prepared resin was packed to a height

of 6 cm in columns consisting of Pasteur pipettes (0.25 x

10 cm) plugged with glass wool. Samples were centrifuged

to remove insoluble material (1 minute @ I6000x g) and

supernatants were loaded on the column. The pellets were

washed with cold PBS, centrifuged, and supernatants

loaded on the column. After loading samples, As(III)

eluted with 7 mL of water. As(V) remained bound to the

resin and the whole column was counted to quantify As(V).

3 9

The amount of arsenic in the insoluble pellet was

determined by gamma counting.

Radiolabelled arsine was not used so several

modifications were made to the above procedure in order

to quantify arsine oxidation products. After As(III) was

eluted, As(V) was eluted from the anion exchange column

with 7mL of 0.5M HCl and insoluble pellets were digested

with a mixture of concentrated HNO3 and 30% hydrogen

peroxide. Arsenic was determined in these samples by

hydride generation. This assay was described by Winski

and Carter (1995). It involves quantitatively reducing

arsenic to arsine with sodium borohydride. Liberated

arsine is trapped in 1 mL of a 0.55% solution of silver

diethyldithiocarbamate in pyridine. Arsenic is then

quantified by absorbance at 510nm.

Arsine was quantified in supernatant of incubations

by mixing ISO^L of supernatant with 2.35mL of 0.55%

silverdiethyldithiocarbamate in pyridine as described

above.

2.1.7. Preparation of cytosol for methylation

experiments. Cytosol was prepared from homogenates by

ultracentrifugation at 105,000 x g for 60 minutes.

Supernatant from this centrifugation step was considered

cytosol.

4 0

2.1.8. Arsenite methyl transferase assay in lung

cytosol. Methylation escperiments were carried out as

described by Zakharyan et al.(1995). Incubations

contained lOpiL 75inM GSH, S.SjiL 70mM DTT, lOpiL 25inM MgCl2,

12.5HL 2M Tris buffer (pH desired for incubation), 200 nL

cytosol, 2nL SAM, 2\iL water, and 10|iL arsenic solution

(25x final concentration). Arsenic concentrations ranged

from 0-250|iM. A control containing protein without

arsenic and a control containing arsenic without protein

were used for each incubation. Incubations were carried

out at pH 8.0 and 37°C unless otherwise noted. Reaction

was stopped by the addition of 10|xL of 40% (w/v) KI, 20

|xL of 15 mg/mL K2Cr04, 750 |iL of HCL, and 750 |iL of

chloroform. Samples were mixed with a vortex mixer for 3

minutes. The organic and aqueous layers were separated

by centrifugation at 2000 x g for 10 minutes. After

extraction, aqueous layer was discarded and replaced with

5HL of 40% KI, 250 piL of water, and 750 of HCl. This

step was repeated 2x. Finally, arsenicals were back

extracted into ImL water. Samples were quantified by

mixing 0.5mL of aqueous phase with lOmL of Universol

scintillation cocktail and lOmL of methanol and counting

with a liquid scintillation counter. Arsenic methylation

4 1

was calculated from dpm of transferred and expressed

as pmol/min/mg protein. Protein was quantified by BCA

protein assay (Pierce Chem. Co.).

2.1.9. Speciation of methylated arsenicals. For

separation of arsenic species from incubations containing

methylated arsenicals, the mixed bed ion exchange method

described by Maiorino and Aposhian (1985) was used.

Briefly this system consists of cation exchange resin

(Dowex 50W-X8, H+ form, 100-200 mesh, J.T. Baker Chem.

Co., Phillipsburg, NJ.)packed on top of anion exchange

resin (AG1-X8, CI- form, 100-200 mesh, Bio-Rad Labs.,

Hercules, CA.). Columns consisted of lOmL disposable

glass pipettes plugged with glass wool. Anion resin was

packed to a height of 4.5 cm; cation resin was packed on

top of anion exchanger to a total height of 16.5 cm.

Columns were washed with 20mL of 0.5N HCl, followed by 25

mL of water, and finally equilibrated in 0.005M

trichloroacetic acid (TCA), pH 2.5. Columns were

sequentially eluted with 28 mL of 0.006M TCA pH 2.5

(ImL/min) , 4 mL of 0.2M TCA (ImL/min) , 28 mL of 1.5M

NH4OH (3mL/min) , and 28 mL of 0.2M TCA (3mL/min) . 2mL

fractions were collected and analyzed for ^H. Arsenite

elutes first, followed by MMA, As(V), and DMA.

4 2

2.1.10. Synthesis of As(SG)3 standard. Sodium

arsenite (Fisher Scientific Co., Fairlavm, NJ) and

reduced glutathione (Sigma Chemical Co., St. Louis, Mo.)

were dissolved in a minimal amount of Milli-Q water in a

1:3.1 molar ratio and allowed to stir for 60 minutes at

room temperature. 5-10 volumes of cold methanol were

used to precipitate product which was collected by

centrifugation. Product was dried by lyophillization.

13 Compound identity was confirmed by C NMR using peak

shifts published by Scott et al, 1993.

2.1.11. Separation and detection of As(SG)3,

Arsenic glutathione complex was separated on a HP1050

HPLC using a Vydac column with standard peptide gradient

of 100% water to 100% acetonitrile containing 0.1%

trif luoroacetic acid (TFA) at a flow rate of imL/min.

The complex was analyzed on a Finnigan 7000TSQ triple

quadropole mass spectrometer using atmospheric pressure

chemical ionization. Standards run by flow injection

were dissolved in water containing 0.1% TFA and injected

into MS at a flow rate of 0.5 mL/min. The carrier was

methanol at a flow rate of 3mL/minute. This methodology

was developed in the analytical core of the Southwest

Environmental Health Science Center, with Dr. Tom

McClure.

4 3

2.1.12. Formation of As(SG)3 in rat lung

homogenates. Male Sprague-Dawley rats were killed by CO2

inhalation. Lungs were perfused with cold saline through

the left ventricle and removed. A 20% (w/v) homogenate

was made in cold saline. Homogenate was mixed with 20mM

arsenite to yield a final As(III) concentration of ImM.

Mixtures were incubated at 37°C. Incubations were

terminated and protein was precipitated by the addition

of TFA to a final concentration of 1% followed by

centrifugation at 16000x g for 10 minutes. Supernatants

were filtered with 0.22|im syringe filters prior to

analysis.

2.1.13. Measurement of Nonprotein Sulfhydryls

(NPSH) . NPSH were determined by the method of Buetler

(1984) using 12% TCA to precipitate proteins. After

centrifugation, deproteinized supernatants were mixed

with 5,5'-dithio-bis(2-nitrobenzoic acid) (DTNB, Ellman's

Reagent) allowing measurement of thiols by absorbance at

412nm. NPSH was determined from a standard curve

produced with GSH. While this assay detects all non­

protein thiols, greater than 90% of cellular NPSH been

reported to be GSH.

4 4

2.1.14. In vitro experiments. Rat lung homogenates

were prepared as described in section 2.1.4. Sodivim

arsenite or arsenate were added to rat lung homogenates

and incubated at 37°C. NPSH concentration was determined

at various times after addition of arsenite or arsenate

by the method of Buetler (1984) with modifications

described by Winski and Carter (1995).

2.1.15. Effect of arsenite complexation on

metaJbolism. 20|iM As(SG)3 was compared to 20(iM sodium

arsenite as a substrate for arsenite methyltransferase in

rat lung cytosol. Methodology was as described in

section 2.1.8.

2.2. Materials and Methods for toxicity studies.

2.2.1. In vitro arsenic toxicity in the lung.

Acute toxicity was investigated in two systems: cultured

cells and lung slices. BEAS-2B cells, an SV-40

transformed human bronchial epithelial cell line, were

obtained from ATCC (Rockville, MD #9609-CRL). Cells were

received at passage 37 and were used between passages 40

and 60. Cells were grown in serum-free modified LHC-9

media (see appendix 1; Lechner and LaVeck, 1985) at 37°C

in a humidified 5% CO2 atmosphere. After incubation,

4 5

toxicity was determined by 2,3-bis[2-methoxy-4-nitro-5-

sulfophenyl]-2H-tetrazolium-5-carboxanilide inner salt

(XTT, X-4251, Sigma Chem. Co., St. Louis) reduction as

described by Roehm et al. (1991) . 96 well plates were

treated with a solution of 10|ig/mL human fibronectin,

30ng/mL Vitrogen 100 (Collagen) , and lOfag/mL bovine serum

albumin dissolved in LHC basal media (VAF coating

solution) for 15 minutes to provide a substrate for cell

growth. 8x10^ cells were plated on VAF coated 96-well

plates in 100|j.1 media. 24 hours after plating, cells were

dosed by adding 50jil of a 3x dosing solution (e.g. for

SO^iM final concentration, cells were treated with 50^x1 of

150HM solution) . 20 hours after dosing, 50|al of XTT

solution was added to each well. XTT solution is made by

dissolving 3 mg XTT in 6 mL of 50°C media, then adding

3.5|il of 30.6mg/mL phenazine methosulfate (PMS, P-5812,

Sigma Chem. Co., St. Louis) dissolved in PBS. Cells were

incubated for 4 more hours and viability was determined by

measuring OD48O ^ Biolinx 2.20 plate reader (Dynatech

Laboratories, Inc.).

Toxicity was also determined by LDH release.

Cells were seeded on T-25 plates and allowed to grow to

approximately 90% confluence prior to treatments. Cells

were treated with 2 mL solutions of arsenic dissolved in

4 6

modified LHC-9 media. To prevent release of arsine gas,

flasks were tightly capped for the first 4 hours of

treatment. After 4 hours, arsine was no longer detectable

in media and caps were loosened to allow gas exchange.

For this assay, media was removed and saved. Cells were

lysed by adding 2 mL of 50mM potassium phosphate buffer

containing 0.5% Triton X-100 to each flask. Media and

cell lysates were kept at 4°C until assay (no longer than

48 hours). LDH activity was determined by LD-L assay kit

(Sigma Chemical Co., St. Louis, MO.). Results were

expressed as % of total LDH activity in media, calculated

as (LDH in media/(LDH in media + LDH in cells)) x 100.

2.2.2. Lung slice experiments. Male Syrian golden

hamsters (120-150g) were killed by CO2 inhalation and

lungs were filled with a 37®C solution of 1.5% gelatin or

0.75% agar in media and placed in ice-cold V-7

preservation buffer until slicing. 8mm tissue cores were

made using an 8mm sharpened stainless steel corer. Slices

were cut on a Brendel/Vitron tissue slicer in cold,

oxygenated V-7 buffer. Slices were ~400jim thick (35-40mg

wet weight/slice) . Slices were kept in cold V-7 buffer

and floated onto teflon/stainless rollers (2/roller).

Rollers were carefully blotted and loaded into 20mL glass

scintillation vials containing 1.7mL Waymouth's MB752/1

4 7

(supplemented with 10% FBS, lOmL/L Fungi-Bact, 50ng/mL

gentamicin, 3.5mg/mL L-glutamine, and 2.4g/L sodium

bicarbonate) (Fisher et al., 1995). Slices were treated

with arsenicals dissolved in Waymouth's media. After 24

hours incubation in dynamic culture incubator, toxicity

was determined by potassium leakage from slices and

histological analysis.

2.2.3. Intracellular potassium assay. Slices were

blotted, weighed, and placed in ImL of ddH20 Slices were

disrupted by sonication (10 seconds, power level 5, Cole-

Parmer Model 4710) . Proteins were precipitated by the

addition of 20|a.L of 70% perchloric acid (PCA) .

Particulate material was removed from samples by

centrifugation at 16,000x g for 10 minutes to obtain clear

supernatants. Potassium concentration in supernatant was

determined on Model 51Ca flame photometer (Bacharach

Instrument Co., Pittsburgh, PA) set on urine potassium.

Potassium concentration was determined from standard curve

of 0-2.OmM solutions of KCl. Potassium concentrations

were normalized to slice wet weight.

2.2.5. Histology of lung slices. After incubation,

slices were floated off rollers into warm Waymouth's media

and transferred with a spatula to 24 well plates

containing warm media. Media was withdrawn with a Pasteur

pipette and replaced with 10% phosphate buffered formalin.

4 8

Slices were fixed overnight and submitted to experimental

pathology service core of the Southwest Environmental

Health Science Center for paraffin embedding, sectioning,

and hematoxylin and eosin (H&E) staining.

2.2.6. Proton Induced X-Ray Emission (PIXE) . The

amount of arsenic in slices was determined by PIXE.

Slices were blotted and placed between layers of mylar

mounted in a sample cup. Holes were placed in the mylar

and samples dried in a dessicator. Samples were then

analyzed by PIXE for arsenic as described by Lowe et al.

(1993) .

2.2.7. Hsp32 induction. BEAS-2B cells were grown to

approximately 90% confluency in 25cm^ flasks (#3055,

Costar Corp., Cambridge, MA.). Cells were treated with

arsenicals in modified LHC-9 media. Preliminary

experiments indicated that maximal induction occurred

around 4 hours after treatment. After 4 hours treatment

with arsenicals, media was removed and cells were rinsed

with 2 mL of sterile PBS-PD (see appendix 1). Cells were

scraped into 0.2mL of sterile PBS-PD and sonicated for 10

seconds to lyse (Model 4710, Cole-Parmer Instr. Co.,

Chicago, IL.). Protein concentration of lysates was

determined by BCA protein assay kit (#23225, Pierce Chem.

Co, Rockford, IL.). If necessary lysates were stored at

4 9

-20®C until analysis. Proteins were first separated by

SDS-PAGE using the method of Laemmli (1970) . lO^g of

lysate protein was loaded for each sample. Standard was

rat recombinant hsp32 (#SPA-895-WB, Stressgen, Victoria,

BC, Canada) . 50ng of standard was used. Gels were 10%

acrylamide with a 37.5:1 ratio of acrylamide:bis-

acrylamide. Gels were cast using Mini-PROTEAN II

apparatus (Bio-Rad Lab., Hercules, CA.) and run at

SOmamps. Western blotting was performed essentially as

described by Burnette (1981). After separation, proteins

were blotted onto PVDF membrane (#162-0181, Bio-Rad Labs,

Hercules, CA.) using Trans-Blot apparatus (Bio-Rad Labs,

Hercules, CA.) in Tris-glycine transfer buffer (pH 8.3,

20% methanol). Conditions for transfer were 60V, 0.21

amps. After blotting, membranes were blocked with 3%

gelatin and probed for hsp32. Primary antibody was

rabbit anti-rat hsp32 (#SPA-895-WB, Stressgen, Victoria,

BC, Canada). A 1:2500 dilution was used for primary

antibody. Secondary antibody was goat anti-rabbit IgG

conjugated to alkaline phosphatase (#A-3812, Sigma Chem.

Co., St. Louis, MO.). A 1:14000 dilution was used for

secondary antibody. After probing, membranes were

developed using Sigma-Fast BCIP/NBT tablets (#B-5655,

Sigma Chem. Co., St. Louis, MO.) as substrate for

alkaline phosphatase.

5 0

2.2.8. DNA single strand break (SSB) determination

by nick translation. Assay was performed as described by

Krause et al. (1993). BEAS-2B cells were grown to

approximately 90% confluency in 25cm^ flasks. Cells were

treated with arsenicals in modified LHC-9 media for up to

24 hours. At timepoints, cells were removed from flasks

with 1% polyvinylpyrrolidine (PVP) in 0.05% Trypsin-EDTA

solution. Cells were collected by centrifugation at 200x

g for 5 minutes and resuspended in 2 mL of sterile PBS-

PD. Cells were diluted 1:1 with Trypan Blue and counted

using a hemocytometer. 2 00nl of cell suspension was put

in well of 96-well vacuum filter plate (0.22um

multiscreen-GV, #MAGVN2210, Millipore Corp., Bedford,

MA.). Cells were washed with isotonic saline (2x), fixed

with 200|il 100% ethanol for 10 minutes, and washed 2x

with saline. Reaction was started by adding 200(il

reaction cocktail (50mM Tris, pH 7.4; 5mM MgCl2; lOmM P-

mercaptoethanol; 50|ig/mL bovine serum albumin, SOpiM each

dATP, dCTP, and dCTP; 3nCi/mL [methyl-3H]dTTP (42

Ci/mmol, l|iCi/|il, Amersham TRK424) ; and 15U/mL DNA

polymerase I. Plate was incubated for 30 minutes at room

temperature. Reaction was stopped by removing reaction

cocktail and washing 5x with isotonic saline containing

2% pyrophosphate. SSB were quantified by removing

5 1

filters from each well with a scalpel, placing in 7mL

scintillation vials with Universol cocktail and

scintillation coianting.

2.2.9 Statistical analysis. N refers to the

number of independent experiments performed. For

cultured cells, experiments were performed on cells of

different passage number. Data are presented as mean ±

standard deviation. ANOVA with Fisher's PLSD post-hoc

test and student's t-tests were used to analyze data

(Statview 4.5,Abacus concepts ), points were considered

significantly different if p values <0.05.

5 2

2.3. Materials and methods for modeling arsenic

metabolism.

2.3.1. The following model of arsenic metabolism

was used for modeling:

( C ) ( B ) ( D ) ( E ) A s ( V ) » A s ( I I I ) • M M A • D M A

\ AsHs ( A )

The following rate equations are derived from this model:

dA/dt=-kAB[A]

dB/dt= IcabCA] + kcB [C]-JcbcC®] ~^^BD [®]

dC/dt= kBc[B]-kcB[C]

dD/dt= kBD[B]+kAD[A]-kDE[D]

dE/dt= koEtE]

ksD and kpg can be described by Michaelis-Menten variables

using the following equations:

kBD=(V„axBD*[B]/(K„BD+[B] ) )

kDE=(V^^DE*[D]/(K^E+[D] ) )

5 3

SIMUSOLV® modeling and simulation software (version 3.0,

Dow Chemical Co., Midland, MI) was used to simulate

arsenic metabolism by the lung. Simulations were run

assuming Img protein in closed incubations. Rate

constants determined in metabolism experiments were

scaled to a per mg protein basis and used as baseline

constants for modeling. If necessary the rate constants

were adjusted to provide the best fit with the data. The

following rate data was used:

reduction: (^cb)

20% homogenate + 1.5mM GSH .000907/min/mg

oxidation: (Kbc)

20% homogenate .00037/min/mg

Conversion of arsine to As(III): (Xab)

20% homogenate .028/min/mg

Conversion of As (III) to MMA: (]Cbd)

lung cytosol Km=5.383 [omol/liter Vmax=.00031nmol/liter/min/mg

Conversion of MHA to DMA: (Icqe)

lung cytosol Km=63.4 |imol/liter Vmax=. 0000384 ^imol/liter/min/mg

5 4

Using this information, the following input file was

created to model the arsenic species produced by

pulmonary metabolism of 100(iM arsenite.

PROGRAM ASH3, Img protein, units are |imol/liter

VARIABLE T ALGORITHM IALG = 2 CINTERVALCINT= 1

CONSTANT KAO=0 .4 CONSTANT KAB=0 .008 CONSTANT KBC=0 .00037 CONSTANT KCB=0 .000907 CONSTANT VM1=0.00031,VM3=0.0000384 CONSTANT KMI=5.383, KM3=63.4 CONSTANT DOSE=IOO CONSTANT TSTOP=120

DYNAMIC

DERIVATIVE

'species A' DADT=-BCAO*A A=lNTEGtDADT,0.0)

'species B' DBDT=KAB*A+KCB*C-KBC*B-(VMI*B/(KM1+B)) B=INTEG(DBDT,DOSE)

'species C DCDT=KBC*B-KCB*C C=INTEG(DCDT,0 .0)

'species D' DDDT=(VMI »B/(KM1+B))-(VM3 *0/(^3+0)) D=INTEG(DDDT,0.0)

'species E' DEDT=(VM3»D/(KM3+D)) E=INTEG(DEDT,0.0)

'total concentration check' TC=A+B+C+D+E

TERMT(T.GE.TSTOP)

5 5

END END END

This is the acsl file for the program. The following

file is the cmd file. Both are needed for simulation.

PREPARET ABCDE set pni=7 start

data T A B C D 0 0 0 0 5 0 . . .0012 15 0 . . .0027 30 0 . 1.4 .0075 60 0 . 1.8 end

PROC pIotA set title='arsine conc' plot A END

PROC plots set title='arsenite conc' plots END

PROC plotC set ti tle=' arsenate conc' plotC END

PROC plotD set titIe='MMA conc' plot D END

PROC plotE set title='DMA conc' plotE END

PROC TC set titIe='total conc' plot TC END

5 7

CHAPTER 3

RESULTS

3.1 Results of metabolism studies.

3.1.1 Inhibition of arsenic methylation. In order to

accurately determine the rates of reduction and oxidation of

arsenic in lung incubations, it is necessary to inhibit

methylation. Because methylation uses arsenite as a

substrate to form MMA, it decreases the amount of arsenite

available for oxidation, changing the equilibrium of the

system. To inhibit methylation, a combination of PAD and

SAH was added to homogenates. The efficacy of this

combination was determined by performing methylation assays

using cytosol prepared from homogenates incubated with PAD

and/or SAH. 100 PAD reduced methylation of 20nM arsenite

by approximately 50%. 1 mM SAH completely inhibited the

methylation of arsenite. The combination of PAD and SAH

also completely inhibited arsenic methylation by lung

preparations (Fig. 2).

3.1.2 Reduction of As(V) by GSH and lung homogenates.

Reduction of As(V) was determined in lung homogenates, using

PAD and SAH to inhibit methylation. Previous work has shown

the importance of thiols, especially glutathione, in

reduction of arsenic. So reduction in lung homogenates was

compared to reduction by l.SmM GSH and lung homogenates

supplemented with 1.5mM GSH. Reduction of 1,10, and lOOiiM

- 0 . 2 -

B D

Figure 2 Inhibition of arsenite methylation by PAD

and SAH in rat lung cytosol. A= 20yM As (III) as

substrate; B= 20pM As(III)+PAD; C= 20viM

As(III)+SAH; D= 20yM As(III)+PAD+SAH. PAD =

lOOnmol/ml periodate oxidized adenosine, added

lOminutes prior to incubation. SAH = Ivimol/ml S-

Adenosyl-homocysteine added just prior to

incubation. Samples were incubated for 15 minutes

at 37°C. Values are mean +/- SD (n=3). * denotes

values significantly different from As(III) alone

(p<0.05), ** denotes values significantly different

fromAs(III) +PAD (p<0.05).

5 9

arsenate was determined. GSH and lung homogenates reduced

arsenic in a time and concentration dependent manner.

Reduction of arsenate proceeded linearly in incubations with

1.5mM GSH and 20% rat lung homogenate supplemented with

1.5mM GSH. Incubations with 20% rat lung homogenate had a

30 minute lag phase, after which reduction was linear to 120

minutes (Fig. 3) . In order to calculate an accurate rate

constant for a reaction, it is necessary to use data from a

linear reaction. The absolute production of As(III)

increased with increasing substrate concentration (Fig 4) ,

however the percentage of As(V) reduced decreased with

increasing concentration (Table l) . This data indicates

that reduction is saturable at high concentrations of As(V).

A rate constant for reduction can be calculated from

this data. Reduction by 1.5mM GSH has a first order rate

constant of 0.0053/minute (r^=0.991); a zero-order rate

constant of 0.00252 |imol/liter/min (r^=0.977). A zero order

rate constant has a better fit to the data than a first

order constant for homogenate incubations. The zero order

rate constant for reduction in lung homogenates is 0.00684

/xmol/liter/min (r^=0.96), the first order rate constant is

0.0104/min (r^=0.93). Supplementing 20% lung homogenates

with 1.5mM GSH increases the zero order constant to 0.0112

/xmol/liter/min (r^=0.886), the first order constant is

0.012/minute (r^=0.73).

0.1 H

0 30 60

time (minutes)

120

GSH homogenate -A— homogenate+GSH

Figure 3 Reduction of lOOyM arsenate [As(V)] by

1. SmM GSH, 20% rat lung homogenate, or 20% rat

lung homogenate supplemented with 1.SmM GSH at

I 37°C. Values are mean +/- SD (n=3-S). * t i denotes values significantly different from GSH

[ (p<O.OS), ** denotes values significantly

different from 20% rat lung homogenate (p<O.OS).

6 1

0.9n

0 . 8 -

fi o o.eq 4-1

— 0.5 -I—( I—I ill. 0.4 -CO

0.3^ f—\ 0 1 0.2^

0.1 -

0

« «*

l.SmM GSH

20% rat lung homogenate

homogenate+l.SmM GSH

*

1 10 100

As (V) concentration (yM)

Figure 4 Reduction of As (V) . Experiments performed

in a volume of 0.5mL at 37°C for 120minutes. Values

are mean ± SD (n=3-5). * denotes values

significantly different from 1.SmM GSH (p<0.05), **

denotes values significantly different from

homogenate {p<0.05).

6 2

As(V) concentration (pH)

system 1 10 100

1.5mM GSH 1.2 ± 0.5 1.1 ± 0.5 0.6 ±0.03

20% lung homogenate 7.3 ± 1.9 7.4 ± 1.6 1.1 ± 0.1

20% homogenate w/ 1.5mM GSH

9.4 ± 2.9 8.9 ± 2.3 1.7 ± 0.2

Table 1. Percent of As(V) reduced after

incubation for 120 minutes at 31°C. Values are mean ±

SD (n=3-5).

3.1.3 Oxidation of As (III) in PBS and lung

homogenates. Oxidation of arsenite was measured in lung

homogenates and compared to that measured in PBS as a

control. Oxidation was not linear over time. Oxidation was

more rapid at 30 minutes than at 120 minutes (Fig. 5). Less

oxidation occurred in homogenates than in PBS at As(III)

concentrations of l^M. However, more oxidation occurred in

homogenates than in PBS with incubations containing IOO/liM

As(III) (Fig 6). The percentage of As(IlI) oxidized

decreased with increasing dose (Table 2).

Rate constants for oxidation were calculated from

this data. A zero order rate constant provided a better fit

6 3

6 0 . 8

O 0.4

40 60 80

time (minutes)

100 120

73 Figure 5 Oxidation of As(III) in 20% rat lung homogenate or PBS. Values are mean± SD (n=3).

* denotes values that are significantly different

from PBS (p<0.05) .

6 4

As(V) concentration (pM)

Figure 6 Oxidation of As(III). Experiments

performed in 0.5mL , incubated at 37°C for 120

minutes. Values are mean ± SD (n=3). * denotes

values significantly different from PBS {p<0.05)

6 5

to the data than a first order. The zero order rate

constant for oxidation was determined to be 0.0094

Mfflol/liter/min (r^=0.912), the first order rate constant is

0.005/min (r^=0.877).

As (113 L) concentration (|iM)

system 1 10 100

PBS 11 ± 0.7 5 ± 2.1 1.4 ± 0.6

20% lung homogenate 4.3 ± 0.1 3.6 ± 0.3 2.3 ± 0.6

Table 2. Percent of As (III) oxidized by PBS and lung

homogenate. Values are mean ± SD {n=3).

3.1.4 Oxidation of AsH3 in PBS and lung homogenates.

Arsine is the fully reduced form of arsenic so it can only

oxidize. Arsine was rapidly lost from solution during the

first 5 minutes of incubation in lung homogenates, after

which time it was depleted at approximately the same rate as

in PBS (Fig 7). The formation of As(III) and As(V) was

detemnined for each incubation. 10.8 ng [144nmol] As(III)

was present in lung homogenate incubations at 5 minutes.

Only 3.4 ng [45.3 nmol] As (III) was present in PBS

incubations at 5 minutes(Fig 8). The formation of arsenite

from arsine is apparently complete within 5 minutes as

arsenite concentrations in these incubations changes very

6 6

80- ,

70-

60-

40-

30-

20-

PBS 10-

lung homogenate

0 5 15 30 60

time (minutes)

Figure 7 Loss of arsine from PBS or 20% lung

homogenate containing ImM arsine incubated at 31°C.

Values are mean +/- SD (n=3-4). * denotes values

significantly differnt from PBS (p<0.05).

12-1

PBS

lung homgenate

4-

2 -

0 + 0

—I—'—'— 10

T—p-i—I—r 20

—r 30

I I I I 40

—n 50

~1 60

time (minutes)

Figure 8 Formation of arsenite [As (III)] in PBS

or 20% rat lung homogenate containing ImM arsine

incubated at 37C. Values are mean +/- SD (n=3).

* denotes values significantly different from PBS

( p < 0 . 0 5 ) .

PBS

lung homogenate

I I I I y f'l i i i i | i i 0 10 20 30 40 50 60

time (minutes)

Figure 9 Formation of arsenate in PBS or 20% rat

lung homogenate containing ImM arsine incubated at

37°C. Values are mean +/-SD (n=3). * denotes

values significantly different from PBS (p<0.05).

6 9

little from 5 to 60 minutes. Arsenate is also formed in

these incubations. The concentration of arsenate also

remains fairly constant with time, actually decreasing at

later time points (Fig. 9) . The apparent first order rate

constant for the loss of arsine from solution was 0.04/min.

The constant for formation of arsenite from arsine that best

fit the data was 0.008/min/mg protein.

Because the compounds used to inhibit methylation could

affect the redox metabolism of arsenic, oxidation of arsine

was also measured in guinea pig lung homogenates. Work by

Healy et al. (1997) showed that guinea pigs do not methylate

arsenite or MMA, so no methylation inhibitors are required.

Arsine disappeared from solution more rapidly in guinea pig

lung homogenate than in PBS (fig 10) . More arsenite is

formed from arsine in guinea pig lung homogenates than in

PBS, indicating more rapid oxidation (Fig. 11). The amount

of arsenate formed from arsine was the same in guinea pig

lung homogenates and PBS (data not shown). These results

are very similar to those obtained from rat lung homogenates

with methylation inhibited. Therefore, use of methylation

inhibitors does not affect the redox reactions of arsenic.

3.1.5. Methylation of arsenite by rat lung cytosol.

Arsenite was methylated by rat lung cytosol. Arsenite

methylation was concentration dependent and saturable (Fig.

12, p. 71). The 15 minute data points from figure 12 were

7 0

80

70-

PBS

lung homogenate c

40-a> c

1 0 -

60 40 50 0 10 20 30

time (minutes)

Figure 10 Loss of arsine from incubations with

PBS or guinea pig lung homogenate. Samples

were incubated at 37°C. Values are mean ± SD

(n=3). * denotes values that are significantly

different from PBS (p<0.05).

7 1

16-1

14-

12-

PBS

lung homogenate

^ 6 -

2 -

40 50 10 20 30 0 60

time (minutes)

Figure 11 Formation of arsenite from arsine

in PBS and guinea pig lung homogenate. Values

are mean ± SD (n=3). * denotes values that

are significantly different from PBS (p<0.05).

7 2

used to create a Lineweaver-Burke plot. Data were converted

to rate (pmol/min/mg) and plotted versus substrate

concentration. This plot indicates that the Kni and Vmav for

arsenite methylation by lung are 5.38 nmol As(III)/liter and

.077 pmol/min/mg respectively (Fig. 13). Because arsenic

methylation is an enzymatic process, the effect of pH on

enzyme activity was determined. For the conversion of

As (III) to MMA, the pH optima was found to be around 8.0

(Fig. 14).

3.1.6. Methylation of arsenate by rat lung cytosol.

Arsenate was methylated much more slowly than arsenite by

rat lung cytosol. Except at high As(V) concentrations,

methylated metabolites only appeared after a lag time.

Unlike arsenite, the rate of arsenate methylation continued

to increase with increasing As concentrations (Fig. 15).

3.1.7. Methylation of arsine by rat lung cytosol.

Methylated arsenicals were detected using arsine as a

substrate. Methylation of arsine was dependent on

concentration, though saturation was not observed (Fig. 16).

The pH dependence of this reaction was investigated and the

optimal pH is around 8.5 (Fig. 17). Because arsine

methylation did not continue to drop at higher pH's, the

possibility of direct chemical methylation was investigated.

No arsenic methylation was detected in incubations of arsine

and SAM, without cytosolic protein (data not shown).

7 3

3.5-1

3-

luM

lOuM

20uM 2 -

cy> e

*—I o e a

1 -

0.5-

5 10 15 20 30 0 25

time (minutes)

Figure 12 Formation of MMA from IpM (•) , lOviM

(•), 20uM (A), and 250iiM (•) As (III) by rat

lung cytosol incubated at 37C. Values are mean

+ /- SD (n=3-5) .

9 0 - ,

1 ' '

o

1 1 1 1 1

o

1 1 1 1 1

o

I 1 1 1 1

o

1 1 1 1 1

o

1 1 1 1 1

o

1 1 1 1 1

o

1 ; 1 1 1

o

1 1 1 1 1

o o

o o o o o o o o o o

o o o o o o o o o o

o o o o o o o o o o

o o o o o o o o o o

*—1 CM fO LT) vo r- CO cr o

1/S

Figure 13 Lineweaver-Burke transformation of

As (III) methylation by rat lung cytosol. Data are

15 minute values from figure 12 (p.73). Values for

1/V are in pmol/min/mg protein.

Vmax = 0.077 pmol/min/mg protein.

Km = 5.383 uM As(III).

7 5

0 . 6 - 1

0.5-

c •H 0.4-^ a> 4-) o

^0.3-1 CT> g

i 0-2-^ o.

0.1 -

6.5 7.5

pH

8.5

Figure 14 Methylation of lOpM arsenite [As(III)]

by rat lung cytosol at various pH. Samples

incubated for 15 minutes at 37°C. Values are mean

± SD (n=3).

7 6

0 . 1 8 luM

0.16-20uM

0.14 - 250uM d •H ^ 0 . 1 2 -o u a. 1 -cn e

0.08 -

0 . 0 6 -6 a

0.04 -

0 . 0 2 -

3lD 20 25 10 15

time (minutes)

Figure 15 Formation of MMA from lioM, 20viM, and

250^M As(V) by rat lung cytosol. Samples

incubated at 37C. Values are mean +/- SD (n=3-5).

7 7

^ lOOuM

g 0.8-

I I 1 ^ I 1 I I

10 15 20

time (minutes)

Figure 16 Formation of MMA from IpM, 20pM/ and

lOOviM AsH^ by rat lung cytosol incubated at 37°C.

Values are mean +/- SD (n=3-4).

7 8

0.5

0.4 -

0) -M 0.3-^ o a CT> e

0 . 2 -o 6 a

0.1 -

7.5

5^

mm

8.5

pH

9.5

Figure 17 Methylation of lOOuM arsine (AsH3) by

rat lung cytosol at various pH. Samples incubated

for 15 minutes at 37°C. Values are mean ± SD (n=3).

7 9

C o •H 4-)

O (0

e a T3

1400 -1

1200 -

1000 -

800 -

600 -

400 -

2 0 0 -

blank

AsH3

ic MMA std

T*-i 1 1 i 1 1 I I I I I . 2 4 6 8 10 12 14 15 18 20 22 24 26 28 30 32 34

elution volume

Figure 18 Separation of arsenic species

formed during 15 minute incubation of lung

cytosol with lOOyM AsH^. Dotted line

represents chromatogram of MMA standard.

8 0

1.8

lOOpM

time (minutes)

Figure 19 Formation of DMA from IpM, lOpM, and

lOOpM MMA by rat lung cytosol. Samples

incubated at 37C. Values are mean +/- SD (n=3).

8 1

Metabolites of arsine were speciated after extraction and

neutralization and found to be largely MMA (Fig. 18).

3.1.8. Formation of DMA from MMA in rat lung cytosol.

DMA was formed from MMA upon incubation with rat lung

cytosol. Formation of DMA increased with increasing MMA

concentration (Fig. 19).

3.1.9. Effect of arsenicals on GSH in rat lung

homogenates. NPSH are depleted rapidly in rat lung

homogenates. Adding As(III) to incubations retarded

depletion of GSH in a concentration dependent manner.

Treatment with similar concentrations of As(V) had little

effect (Fig. 20). Comparison of As(SG)3 to GSH in NPSH

assay showed that assay conditions dissociated the complex

and resulted in approximately 3 moles reduced GSH per mole

of complex analyzed (Table 3).

8 2

AS(SG)3 concentration assayed NPSH concentration

(HM) (HM)

1 3.3 ± 10*

10 38 ± 3.14

50 152 ± 5

100 304 ± 24

Table 3. NPSH content of aqueous solutions of As(80)3.

NPSH was measured by Ellman's reagent as described by

Buetler (1984). Values are mean ± SD (n=3). * denotes

samples that were near the limit of detection, causing high

SD.

300 -1

control

250 0.25iiiM As(III)

2inM As (III) 4-1

^ 200 -Z3 o c •H

150 -

2.5inM As (V)

K CO Oj Z 1 0 0 -

o

50 -

2.5 3 2 1.5 0 0.5 1

Time (hours)

Figure 20 Non-protein sulfhydryls in rat lung

homogenate treated with arsenicals. 1.5ml samples

were incubated at 31°C for times indicated.

Values are mean ± SD {n=3). * denotes values

significantly different from control (p<0.05).

8 4

3.1.10. Formation of As(SG}j in rat lung. As(30)3

standards were isolated and detected using LC-MS. As(SG)3

is more stable at low pH. By using 0.1% TFA in the mobile

phase, the complex could be chromatographed without

dissociating. The three major peaks in the MS spectra had

m/2 ratios of 994, 498, and 332; corresponding to the

singly, doubly, and triply charged forms of the complex.

MS-MS was performed on the 994 ion, resulting in major peaks

with m/z ratios of 687, 558, 380, and 308 (Fig. 21 and 22).

Using this system, As(SG)3 was detected in rat lung

homogenate after incubations with sodivm arsenite for 15

minutes (Fig. 23).

Sodium arsenate was shown to be reduced and then form

complexes in solutions containing GSH by Scott et al.

(1993). To determine whether similar reactions occur in

biological systems, As(SG)3 formation by rat lung

homogenates was investigated using ImM sodium arsenate or

ImM sodium arsenite as a substrate. In these incubations,

AS(SG)3 was detected after 5 minutes incubation with ImM

sodium arsenite. The concentration of the complex decreased

steadily with time. At 60 minutes, only 11.5% of the

initial concentration of AS(SG)3 remained. In incubations

containing ImM sodium arsenate, a small amount of AS(SG)3

was detected at 5 minutes, but not at any other time (Fig.

24) .

IkZO 100-1

8 0 -

6 0 -

40

20

332 498 E.OS B 85

9M

533

1007 1065 1338 1402 B07 852 868 edo

Figure 21 Mass spectrum of As(SG)3 standard. Sample inlet by

flow injection. Major peaks correspond to singly (994), doubly

(498), and triply (332) charged parent ions.

CD cn

8 6

3

179

130 j 205 251

! , 1 1 1"^

38

38

36^

1

6E

0

557 1

JSl 616

1 ! 1

7

689

366 920 975 849 i I95I 1

—1 1 1 -1 1 ,

SSB

d ' 2(lo 4^ 0 6(io ' ado ' lo'co

Figure 22 Daughter ion spectrum from 994 fragment of As(SG)3. Samples ionized by APCI on Finnigan TSQ9000 mass spectrometer.

•/St74>7S ac 7

100 -

0 -'

lOO <

BS30<>30I

^ -m/xtA99 424 7197

i 433 7:p«

I. .l.tJ • . • - J i-

••03 943

1331 30^31

fki-rtr-i TirV- fMln n -id-tfi i. ri: jlJ

1^03 (47

. ••04 r3.74f

. B^3 r3.«M

41f 7:P2 . B^OO 15.710

L I ' •>' ' I'liOflMiS'' I' f'M" 1-* (* Y"[>' 1 -n I**-! -t'l A Soo 1000 isoo

r B^OS lf3

B.

CO '

40

^5^0 4^

400

. 8«-04 4.04

Sit I

-n-' 500

M I

s«o I S^C 43J 4C1 4^4

I <1 V I » <00 700 •00 900

Figure 23 A) Selected ion chromatogram of As(SG)3 standard run on Vydac coliamn with standard peptide gradient. B) Mass spectrum of 7.07 minute peak from A.

8 8

•/Z>74>7(

ftA .. A A-AA/U^ J\«AA.A^aA<SAAJ1A7\

at 7

i t*03

S.345

•/S<30<>30t 32J 5»?2

JL r-•»04 943

•/St49t 433 Tspf

i i-

t^04 33t

t^03 r3>3SO

USSR 419 7tpa

SK 7

. s*oo 13.<43

£ A».-A-A/W|N,A-p-,-^-^AA^-/> .A -ftA / K^as •73

200

>•04

^0 icp i l l . . - I I

Figure 23 C) Selected ion chromatogram of supernatant of rat lung homogenate incubated with ImM As(III) for 15 minutes at 37°C run on Vydac column with standard peptide gradient. D) Mass spectrum of 7,06 minute peak from C.

8 9

4000

3500 As (III)

3000 As (V)

standard ^ 2500

\ 2000 e

o 1500 -S

1 0 0 0 -

500 -

60 50 30 40 20 0 10

tme (minutes)

Figure 24 As {SG) in J^at lung homogenates

incubated at 37°C with ImM As(III) or ImM As(V).

Standard is lOuM As(SG)2 PBS incubated at 37°C.

Values are mean +/- range (n=2).

9 0

45000

40000 -

35000 -

-o 30000 -(U

^ 25000 in C

20000 -

-a 15000

10000 -

5000 -

As(III)

As (SG) 3

1 ' ' ' ' I ' ' ' ' I

10 15 20

time (minutes)

25 30

Figure 25 As(SG)3 and As (III) as substrate for

arsenite methylatransferase. 200^1 rat lung

cytosol was incubated with 20viM substrate at 37°C.

Values are mean ± SD (n=3).

9 1

3.1.11. AS(SG)2 as substrate for arsenite

methyltransf erase. As (86)3 and As (III) were compared as

substrates for arsenic methyltransferase in rat lung

cytosol. 20|IM AS(SG)3 and As (III) were methylated at equal

rates in this system (Fig. 25) indicating that both forms of

arsenic were equivalent substrates for the arsenite

methyltransferase enzyme.

3.2 Results of toxicity studies

3.2.1. Effect of arsenicals on BEAS-2B viability.

Acute arsenic toxicity in the lung depends on arsenic

species. In BEAS-2B cells cultured with arsenic for 24

hours, As (III) had an LC50 of 40|iM; As(V) had an LC50 of

12OHM; MMA had an LC50 of 2mM; and DMA had an LC50 of lOmM.

Arsine had an LC50 of 750|aM in the XTT assay (Fig. 26) .

Because arsine may have volatilized from 96-well plates,

studies were also run in sealed flasks using LDH release as

a measure of toxicity. ImM arsine did not significantly

increase LDH release for at least 7 hours in BEAS-2B cells.

However ImM arsenite treatment caused significant increases

in LDH release by 5 hours (Fig. 27).

3.2.2. Toxicity of arsenicals in hamster lung slices.

Toxicity in hamster lung slices also depends on the chemical

form of arsenic. Incubation with lOOjiM As (III) for 24 hours

I

120- , As(III)

As (V) 100

MMA >1 4-) •H

80-

(0

DMA

AsH3 •H >

I—I o M -M

60-

a o u

40-O

dP

20-

0 . 0 0 1 0 . 0 1 0 . 1 1 10 100 As concentration (mM)

Figure 26 Viability of BEAS-2B cells treated with

arsenicals for 24 hours (as determined by XTT assay).

Values are mean ± SO (n=3). vo

16-,

14-

12-

control

0 ) 1 0 -As (III)

As (V)

AsH3

6 -

<*>

2 -

time (hours)

Figure 27 LDH release from BEAS-2B cells treated

with litiM arsenicals. Values are mean ± SD {n=3) ,

* denotes values significantly different from

control.

9 4

4->

2 4-> CD

0.04

0.035

0.03 -

0.025 -

^ 0 . 0 2

H-

o e 3

0.015 -

0.01 -

0.005 -

1 I 1 I

control As (III) As (V) AsH3

Figure 28 Potassium leakage from agar filled

hamster lung slices treated with lOOpM arsenicals

J for 24 hours at 37°C in Waymouth's media +10% FBS.

' Values are mean ± SD {n=3-5). * denotes values

[ significantly different from control (p<0,05). I 4 I ]

9 5

produced a 29% decrease in intracellular potassium; 100|iM

arsine produced a 47% decrease. lOOuM As(V) had no effect

on intracellular potassi\im (Fig. 28) .

3.2.3. Arsenic content of slices. The amount of

arsenic accumulated by lung slices was determined by PIXE in

order to correlate toxicity with arsenic content. Both

As(III) and As(V) treated slices contained about 215 ng

As/cm2. AsH3 treated slices contained slightly more

arsenic, 256 ng As/cm2 (Fig. 29).

3.2.4. Histology of arsenic treated lung slices.

Control slices showed signs of alveolar edema, but airway

epithelia were largely intact. Arsenite treated slices did

not exhibit edema of the alveolar walls, but alveolar cells

were pyknotic. Airway epithelial cells were vacuolated and

pyknotic. In arsine treated slices, alveolar cells are

swollen, vacuolated, and pyknotic. The airway epithelial

cells have been sloughed in arsine treated slices(Fig. 30).

3.2.5. Effect of arsenicals on Hsp32 induction.

Hsp32 levels were determined in BEAS-2B cells treated with

arsenicals for 4 hours. Arsenite, arsine, and arsenate

induced hsp32 expression (Fig. 31) . Induction of hsp32

was dose-dependent, with maximal induction by arsenite

between 10 and 30^M; maximal induction by arsine occurred

between 100-300|iM; maximal induction by arsenate also

9 6

300-1

250-

200-

c\i

" 150-

100-

50-

AsH3 As(III) As (V) control

Figure 29 Arsenic accumulation in agar filled hamster

lung slices treated with lOOpM arsenicals for 24 hours

in Waymouth's media + 10% FBS. Slices incubated at

37°C with 95%02. Values are mean ± SD {n=3). *

denotes values significantly different from control

(p<0.05).

Figure 30 Histology of agar-filled hamster lung slices incubated with lOOjiM arsenicals for 24 hours. A) Control B) As(III) C) ASH3

A. Lane 123456789 10

Figure 31 Induction of hsp32 (heme oxygenase 1) in BEAS-2B cells treated with arsenicals for four hours. Western blot probed witgh monoclonal antibody to rat hsp32 (cross reacts with human hsp32) . Treatments were as follows A) Lane 1,

ImM DMA; 2, ImM ASH3; 3, 300nM ASH3; 4, 100|iM ASH3; 5, 30tAM ASH3; 6, 100|iM As(III) ; 1, 30nMAs(III); 8, lOjiMAs(III); 9, lUM As(III); 10, 50ng standard. B) Lane 1, control; 2, 10|iM As(III); 3, 50UMAs(V); 4, lOOnMAs(V); 5, 20piMAs(V).

1 0 0

60000-1

40000 i/j 4-)

c 3 O U 4-1 0) c 20000

treatment

Figure 32 Quantitation of blot in figure 31.

Lanes analyzed by densitometry. Induction of

hsp32 in BEAS-2B cells treated with arsenicals

for 4 hours. 20vig protein loaded per lane. Net

counts were determined by densitometry, using

background subtraction for each lane.

1 0 1

22000 n

20000 -

(f)

o

§ 18000 o m

E a. XJ

/

2000-

i i 24

i n 'A

neg control

pos control

5[iM As(lll)

1mM arsine

1mM DMA

100|JM MMA

.. K

"s;

i % I 12

time (hours)

Rgure 33 Single strand DNA breaks in BEAS-2B cells treated with arsenicals. Positive control was 400uMMNNG for 20 hours. Values are mean ± SD (n=4-6). 'represents treatments different than negative control (p< 0.05).

102

occurred between 100-3OOnM. Arsine produced greater

induction of hsp32 than arsenite (Fig. 32).

3.2.6. Results for single strand breaks. BEAS-2B

cells were treated with equitoxic concentrations of

arsenicals for up to 24 hours. 400|iM MNNG treatment for 20

hours was used as a positive control. MNNG treatment

produced a 10-fold increase in SSB over 24 hour negative

control, increasing from 2,000 to about 20,000 dpm/150,000

cells. No arsenical produced significant increases in SSB

with 4 hour treatment. ImM DMA induced a significant

increase in SSB after 12 hours of treatment. SSB returned

to control levels in all arsenical treatments by 24 hours

(Fig. 33) . To determine if metabolism of MMA to DMA could

cause SSB, cells were incubated for 12 hours with

concentrations of MMA up to ImM. Even though these doses of

MMA were toxic, no increases in SSB were observed.

3.3 Results of modeling studies

3.3.1. Modeling pulmonary arsenic metabolism. A model

of pulmonary metabolism of arsenic was constructed using the

data collected in metabolism studies. The accuracy of the

model was tested by comparing observed and calculated

concentrations of metabolites produced by various

arsenicals. Using lOO^M arsenate as the initial condition.

103

B

Figure 34 Simulated metabolism of lOOuM As(V). Values

on y axis are in x axis are minutes. Open boxes represent observed values, solid lines represent calculated values. A) formation of As(III) B) loss of As{V) C) formation of MMA D) formation of DMA.

104

B

ISO too «0

(_l • o

120 M 20

A B

O

Ui

O

•0 •0 %0

o

o •

Q

120 «0 100

Figure 35 simulated metabolism of lOOjaM As (III).

[ Values on y axis are in nM, x axis are minutes. Open boxes [ represent observed values, solid lines represent calculated I values. A) loss of As(III) B) formation of As(V) C)

formation of MMA D) formation of DMA.

105

•r«in« conc cone

B HHP cor»e

C D OHfl eonc

m b

Ul

•

Q

120 BO too •0

E Figure 36 Simulated metabolism of imM ASH3. Values on

y axis are in (iM, x axis values are minutes. Open boxes represent observed values, solid lines represent calculated values. A) loss of arsine B) formation of As(III) C) formation of As(V) D) formation of MMA E) formation of DMA.

106

observed and calculated concentrations of As (III) and MMA

were very similar (Fig. 34) . Using lOOjiM arsenite as the

initial condition, observed and calculated concentrations of

As(V) and MMA were very similar (Pig. 35).

The metabolism of arsine presented some interesting

challenges. The oxidation is so rapid that rates for loss

of arsine and formation of arsenite could only approximate

those observed. Using these approximations, it was possible

to use the model to determine if arsine must be oxidized to

arsenite before being metabolized to arsenate or MMA. Using

the parameters for oxidation of arsenite and methylation of

arsenite, but using ImM arsine as the initial condition, the

observed concentrations of arsenate and MMA are higher than

calculated values (Fig. 36).

3.3.2. Correlation of metabolism and toxicity data.

Arsenate and arsine are metabolized to arsenite by the lung.

Because arsenite causes cell death and hsp32 induction at

lower doses than either arsenate or arsine, it is possible

that arsenite is the "active" form of arsenic in all cases.

DMA was the only arsenical that produced SSB, so it must be

the "active" form as it should not be further metabolized.

It may be possible to explain the results of the toxicity

studies by modeling the concentration of the putative

"active" form of arsenic [arsenite or DMA] present after

treatment with each arsenical. The arsenic species

107

distribution in the toxicity assays were modeled. The

concentration of arsenite or DMA predicted by this

simulation are presented in Table 4.

Induction of Hsp32

Initial conditions As(III) concentration

at 4 hours

IHM As(III)

lOiiM As (III) 9.2HM'^

lOjxM As(V) 0.6|IM'

20HM As(V)

lOOliM AS(V) 6|iM

lOuM ASH3

30011M ASH3

Table 4a Correlation of predicted arsenite

concentration at 4 hours with hsp32 induction (4 hours is

the time of maximal induction according to preliminary

experiments) . ^ denotes minimtim concentration at which

induction was observed. ^ denotes conditions at which

maximal induction was observed. denotes conditions under

which induction was not observed.

108

LC50 in BEAS-2B cells treated for 24 hours

Initial conditions As(III) concentration

at 24 hours

40^M As(III) 30HM

120|iM As(V) 31nM

750HM ASH3 115HM

Table 4b Correlation of predicted arsenite

concentration at 24 hours with effects on cell viability.

Induction of single stranded DNA breaks

Initial conditions DMA concentration at 12 hours

ImM MMA 25|iM

Table 4c Correlation of predicted DMA concentration at

12 hours with induction of single strand DNA breaks.

109

CHAPTER 4

DISCUSSION

My research had four goals: 1) to determine the

metabolism of arsenic by the lung, 2) to determine the

effects of various forms of arsenic on the lung, 3) to

develop a mathematical model of arsenic metabolism in the

lung, and 4) to determine whether arsine produces effects by

oxidation to arsenite. I hypothesize that it is possible to

correlate the effects of arsenicals with the concentrations

of arsenic produced by metabolism.

4.1 Metabolism Studies

To understand the effects observed after inhalation of

arsenic, the species of arsenic actually present at a given

time must be known. The distribution of arsenic species

after exposure is a result of the chemical and biochemical

reactions that arsenic undergoes in the lung. The sum total

of these reactions is being called metabolism, as all are

reactions that change the form of arsenic present in the

tissue. Previous metabolism studies have largely been

performed in vivo. Animals were dosed with one form of

arsenic and the species of arsenic present in the urine were

determined and used to assess whole body arsenic metabolism.

This provides a panorama of arsenic metabolism but does not

110

provide much information about the metabolism of arsenic in

individual tissues, especially target tissues.

The lung is the site of exposure to inhaled arsenic, as

well as the target organ. For this reason, it is probable

that other organs are not involved in the metabolism of

arsenic as it affects toxicity in the lung. Inhalation of

arsine is an exception to this model. Arsine is a gas and

exposure is almost exclusively by inhalation, but arsine is

a potent hemolytic agent causing little pulmonary toxicity

in vivo. In this case, metabolism of arsine by the lung

will act as a first pass effect, controlling the form of

arsenic absorbed by the blood.

The only previous study on the metabolism of arsenic by

the lung focused on methylation and indicated that the lung

was deficient in arsenic methylation (Georis et al., 1990).

This could have serious effects on the susceptibility of the

lung to arsenic toxicity. In order to understand toxicity,

the species of arsenic actually present after exposure to

various arsenicals must be known. If the rate of each

metabolic step is known, the arsenic species distribution

for any situation can be calculated. The first step in

creating this model was to quantify the rate of each step in

pulmonary arsenic metabolism.

4.1.1 Redox reactions of arsenic in the lung. Arsenic

metabolism is a complex process that involves competing

redox reactions and methylation reactions. To accurately

Il l

quantify any given step, competing processes must be

eliminated. This is especially important for oxidation, as

methylation of As(III) is an oxidative process that competes

with oxidation to As(V). Because S-Adenosyl-L-Methionine

(SAM) is the methyl donor for arsenic methylation, periodate

oxidized adenosine (PAD) has been used previously to

competitively inhibit methylation in vivo (Marafante and

Vahter, 1984; Marafante et al., 1985;). PAD inhibits S-

Adenosyl-homocysteine (SAH) hydrolase, which metabolizes SAH

to adenosine and homocysteine, causing an increase in SAH

concentrations. SAH is a competitive inhibitor of most

methyltransferases. In our in vitro system, PAD alone only

reduced methyltransferase activity by 50%, however, addition

of exogenous SAH in combination with PAD, provided complete

inhibition of arsenic methyltransferase activity. Use of

this system allowed accurate measurement of the redox

reactions of arsenic in the same species in which

methylation was measured.

Arsenate is reduced to arsenite by the lung. More

arsenic was reduced by lung homogenates than by GSH alone

indicating that there are other mechanisms besides chemical

reduction occurring in lung cells. The percentage of

arsenic reduced decreased at higher arsenic concentrations,

indicating that the processes involved in arsenic reduction

may be saturable. As seen in table 1, saturation does not

112

occur until As(V) concentrations exceeded lOuM, a value

unlikely to be exceeded in vivo, so saturation of reduction

is probably not a concern.

Although numerous studies have observed that there is

reduction of As(V) in vivo, there have been very few

previous studies that measured the reduction of As(V) by a

specific organ. Marafante et al. (1985) concluded that most

in vivo reduction occurs in blood but did not measure it.

Ginsburg (1965) measured the appearance of arsenite in the

urine and perfusate after infusing arsenate into the renal

artery. Using this methodology he calculated that llOnmol

As(III)/min was formed by the kidney. Plotting reabsorption

of As(V) vs. secretion of As (III) he obtained a slope of

0.2nmol As(III) secreted/ nmol As(V) reabsorbed. Ginsburg

postulated that equal amounts of As(III) were secreted into

urine and plasma, so actually a slope of 0.4nmol

As(III)/nmol As(V) reabsorbed is obtained. Based on this

data, Mann et al. (1996) calculated a first-order rate

constant for kidney reduction of 1.75/hr. Based on in vivo

studies by Marafante et al. (1985), a first order rate

constant for whole body reduction of 1.37/hr was obtained.

By comparison, the first order rate constant for reduction

of As(V) by 1.5mM GSH is 0.32/hr. Reduction by 20% rat lung

homogenate has a first order rate constant of 0.627/hr.

Adding l.5mM GSH to 20% lung homogenates increases the first

113

order rate constant to 0.739/hr. These data suggest that

the lung reduces arsenate at about half the rate of the

whole body or kidney, and is certainly capable of reducing a

significant portion of arsenic absorbed during inhalation

exposure.

As (III) oxidizes to As(V) in lung homogenates. The

extent of oxidation was dependent on As(III) concentration

(Fig. 6) . This is actually a measure of net oxidation, as

reduction is occurring simultaneously with oxidation. With

IjoM As (III), less oxidation occurred in homogenates than in

PBS. With lOOjiM As(III), more oxidation occurred in

homogenates than PBS. This probably occurs because the

extent of reduction slows as arsenate concentrations

increase, leading to increased net formation of As(V) at

higher concentrations.

No studies have measured the oxidation of arsenic in

vitro biological systems, so comparing the lung to other

organs is difficult. Based on in vivo studies, Mann et al.

(1996) calculated a first order rate constant for oxidation

of 1.83/hr. By comparison, arsenic oxidizes in the lung at

an apparent first order rate constant of 0.30/hr. The fact

that this constant is so low compared to the whole body

constant is surprising given the high O2 content of the

lung. Conducting these experiments in lung homogenates may

have affected the oxidation of arsenic as oxygen tension was

114

probably somewhat lower than it is in the lung in vivo. The

in vivo study used to calculate a whole-body oxidation rate

constant used urinary arsenic profiles. It is possible that

a considerable eumount of arsenic oxidized once in urine,

especially if urinary pH was above 7.

Arsine is the fully reduced form of arsenic so the only

pathway for its metabolism is oxidation. Arsine disappears

from solution more rapidly in lung homogenates than in PBS.

This could be due to several processes. Arsine is a gas, so

some of the loss of arsine from solution is due to

volatilization. Arsine may volatilize more rapidly from

lung homogenates than from PBS. Arsine is highly

lipophillic, so arsine may be dissolved in membrane

fragments in homogenates. These fragments are removed prior

to arsine analysis, so arsine dissolved in these fragments

would be lost. Arsine may also be oxidized more rapidly in

lung homogenate than in PBS. In this case, more oxidized

arsenic should be found in lung homogenates than in PBS.

Indeed, there are 17 /xg less arsenic (as arsine) in lung

homogenate incubations of arsine after 5 minutes incubation

than in PBS incubations. Of these 17 |ig, 9 more pig are

present as arsenite in lung homogenate incubations than in

PBS. These results confirm that arsine oxidizes more

rapidly in lung homogenate than PBS.

115

To confirm our model, and ensure that the methylation

inhibitors were not affecting redox reactions, arsine

oxidation was also measiured in guinea pigs. Guinea pigs are

deficient in arsenic methyltransferase activity (Healy et

al., 1997) and do not methylate arsenite. Therefore, redox

reactions can be measured directly. Arsine oxidation by

guinea pig lung homogenates was essentially the same as that

by rat lung homogenates with methylation inhibited. This

would indicate that the methylation inibitors were not

affecting our results.

The oxidation of arsine occurs very rapidly. The lines

in figure 7 have different slopes to 5 minutes, but are

parallel after that. Also, the As(III) and As(V)

concentrations are relatively stable, indicating that active

oxidation of arsine by lung only occurred during the first

five minutes. Why this process should stop is not clear,

but may involve exhaustion of some component of the

reaction. It is possible that there is an enzyme

responsible for oxidation of arsenic in the lung, similar to

that reported by Osborne and Ehrlich (1976) in bacteria.

These bacteria oxidize arsenic by transferring electrons to

oxygen via an oxidoreductase. It is also possible that a

reaction between arsine and solution oxygen is catalyzed in

lung homogenates. In this case, once solution oxygen is

depleted, oxidation would stop.

116

Another interesting point is the loss of As(V) from

incubations of arsine and PBS. Because PBS has no reductive

ability, this implies a reaction between arsine and As(V).

This reaction could proceed as depicted below;

AsH3 + As(V) + O2 > 2As(III) + 2H2O.

The production of As (III) and As(V) does not account

for all of the arsenic lost from incubations of arsine. As

mentioned, volatilization may account for some of the lost

arsenic, but it is likely that there were some species of

arsenic, such as As(0) and AS2H4 (arsine dihydride), present

in the incubations that were not measured. Analytical

techniques that are capable of detecting these other arsenic

species need to be developed before arsine metabolism can be

completely determined.

Arsine was not detected in incubations of As(III) with

lung homogenate. This is not surprising as the

thermodynamics of the reaction; As(III) + 6e- > AsH3 are

very unfavorable (E° = -1.22V, Weast, 1976) and a strong

reductant, such as sodium borohydride is required.

These data were collected in lung homogenates buffered

at neutral pH and under ambient air. However, redox

reactions of arsenic are affected by pH and oxygen content.

As pH decreases the reduction potential of As(V) decreases.

This is shown in the following graph.

117

0.2-

0.1-

0-

-0.1-

-0.2-

•n 0 2 4 6 8 10 12 14

PH

Figtire 34 Effect of pH on the formal reduction

potential of arsenate.

Thus, in areas of low pH, such as the stomach or

lysosomes. As(III) will be the favored species of arsenic.

The interior of mitochondria is at a higher pH than the

cytosol, favoring As(V) species. This can have a

significant impact on toxicity and should be addressed when

predicting arsenic species distributions in various regions

of the cell or body.

Because these experiments were carried out in

homogenates, uptake was not a factor. The data is

indicative of the ability of the cell to metabolize

intracellular arsenic, but does not account for the uptake

118

of arsenic which may be limiting. Work by Winski and Carter

(1995) showed that uptaUce of arsenate was more rapid than

its reduction in erythrocytes and, thus, was not the

limiting step in arsenate metabolism. However Lerman et al.

(1983) found virtually no uptake or further metabolism of

arsenate by cultured hepatocytes. The discrepancies in

these two studies indicate that arsenate uptake is tissue

specific and should be studied further. Arsine is

lipophillic and should diffuse rapidly into the cell. Free

arsenite is uncharged at physiological pH and also readily

enters cells. Several studies have demonstrated that

arsenite freely diffuses across biological membranes

(Ginsburg, 1965; Huang and Lee, 1996). Therefore,

metabolic studies in homogenates should approximate

metabolism by intact cells fairly well. In fact, it is

likely that whole cells will reduce arsenic more rapidly

because the microenvironment of the cell remains intact.

4.1.2. Methylation of arsenite and MMA in the lung.

The other type of reaction involved in arsenic metabolism is

methylation. Arsenite is methylated by rat lung cytosol to

MMA and MMA is further methylated to DMA. The extent of

methylation is dependent on the chemical form of arsenic

present. Arsenite is methylated most rapidly, followed by

arsine and arsenate. This fits the oxidative methylation

theory of Challenger (1945). Methylation of arsenite occurs

by enzyme mediated transfer of methyl groups from S-

119

adenosyl-L-methionine to arsenic(III) . This is an oxidative

methylation, so arsenic must be in the +(III) state

initially and be oxidized to the +(V) state during

methylation as described by Challenger (1945). As(V)

species, including MMA, must be reduced prior to

methylation. For this reason, methylation of As(V) is much

slower than for As(III). Thompson (1993) proposed a pathway

for arsenic methylation that required GSH to reduce As(V)

species to As(III) species but not for the actual

methylation reaction. He proposes that there are different

methyltransferases for each methylation step and that a

dithiol cofactor is required for methyltransferase activity.

Work by Zakharyan et al (1995) showed that arsenite and MMA

methyltransferase activity copurify and appear as a single

band on SDS-PAGE. This indicates that both activities are

part of a single protein or protein complex.

The optimal pH for arsenic methylation by rat lung

cytosol was around pH 8.0. The optimal pH observed for

rabbit liver cytosol was reported to be 6.8 by Zakharyan et

al (1995) . This group has purified the arsenite

methylatransferase enzyme from rabbit liver 2000-fold. The

optimal pH for purified enzyme is 8.0. The difference in

optimal pH between cytosol and purified enzyme is proposed

to be due to the presence of inhibitors in the cytosol. The

fact that the optimal pH for methylation of arsenic by rat

lung cytosol so closely matches that of the purified rabbit

120

liver enzyme may indicate that the inhibitors present in the

liver are not present in the lung, or not present in the

rat. The similarity in optimal pH also suggests that the

same enzyme is active in rat lung and rabbit liver.

In lung cytosol, arsine was methylated to MMA. How

this occurs is not clear. Arsine is oxidized rapidly in

lung homogenates, forming significant amounts of arsenite.

This arsenite could then be methylated. It is also possible

that arsine is methylated without oxidizing to arsenite.

This would presumably form monomethylarsine which

subsequently is oxidized to MMA. The methylation of arsine

does not occur without cytosolic protein in the incubation,

indicating that enzymes in the cytosol are responsible for

the methylation of arsine. The optimal pH for this reaction

was determined to be around 8.5. This is similar to that

observed for methylation of As(III). The slight increase in

optimal pH may reflect the need for oxidation of arsine to

arsenite prior to methylation.

Arsenite methylation in lung cytosol follows Michaelis-

Menten kinetics, and is saturable at high concentrations of

As(III). There has been considerable debate on whether

saturation of arsenic methylation is observed in humans. A

study by Hopenhayn-Rich et al. (1996) found no evidence for

exposure based threshold on human arsenic methylation. They

did find that other factors, such as cigarette smoking and

mercury intake can have a significant effect on arsenic

121

methylation. If arsenic methylation is saturable in humans,

it could contribute to a threshold dose for arsenic induced

cancer.

Rat lung cytosol forms both MMA and DMA from inorganic

arsenic in in vitro incubations. Work by Georis et al.

(1990) indicated that rat lung slices produced only 30% of

the methylated arsenic produced by liver slices. Our work

indicates that rat lung is capable of methylating As at

approximately the same rate as rat liver (Barber et al.,

1995). This difference is probably due to differences in

assay systems. In slices, tissue architecture is maintained

and arsenic must get into the cell before it is methylated.

In cytosol incubations, arsenic has direct access to enzymes

involved in methylation, speeding the reaction. Our work

indicates that the lung is capable of methylating arsenic as

efficiently as the liver. However, based on the work of

Georis et al. (1990), there may be different rates of

uptake.

It is accepted that methylation of As(V) requires

reduction to As(III) prior to methylation. Using this

reasoning, MMA is believed to be reduced to MMAs(III) prior

to methylation to DMA. MMA and DMA are reduced to MMAs(III)

and DMAs(III) by thiols (Cullen et al., 1984). Comparison

of the rate of As(V) methylation to MMA methylation shows

that MMA is methylated faster than inorganic As(V). There

are two steps involved in methylation of As(V) species.

122

reduction and methylation. Because MMA is methylated more

rapidly than As (V), it must proceed through these two steps

faster. There are several possible explanations. One

possibility is that the enzyme responsible for methylating

MMA works faster. This is not the case, as Vmax is higher

and Km is lower for inorganic arsenic than MMA. Therefore,

MMA must be reduced faster than As(V), providing more +(III)

substrate for the methylating enzyme. This may occur as the

methyl groups may decrease the reduction potential of the

arsenic.

The formation of methylated As(III) species is an

overlooked area of arsenic toxicology. The body clearly is

capable of reducing arsenicals. Cullen et al. (1984)

detected both MMAs(III) and DMAs(III) in solutions

containing thiols. Although As(III) is found in the urine

of exposed individuals, MMAs(III) and DMAs (III) have never

been isolated from biological systems. Based on the

oxidative methylation hypothesis, MMAs(III) must form during

the formation of DMA. It is possible that MMAs(III) only

forms during the methylation process, perhaps through the

activity of an intrinsic reductase activity of the

methyltransferase. However, it is more likely that

MMAs(III) and DMAs(III) are formed by the same pathways that

reduce inorganic arsenic and just have not been isolated

from biological samples. Evidence for the formation of

DMAs(III) by GSH is found in work by Ochi et al. (1996).

123

Treatment of HL-60 cells with arsenicals caused apoptosis.

Reducing GSH levels in the cells prior to treatment with

arsenicals increased apoptosis for all arsenicals except

DMA., where it decreased. These data indicate that GSH is

protective against inorganic arsenic and MMA toxicity, but

potentiates DMA toxicity. This probably occurs as a result

of GSH reducing DMA to DMAs (III) which is more toxic than

DMA.

4.1.3 Complexatlon of arsenite with glutathione. The

last process that changes the form of a metal in the body is

complexation. Many metals form complexes with physiological

ligands. These complexes are the species that actually

exist within the body and determine metal transport and

toxicity. Knowing the species present in the blood and

inside the cell allows more accurate interpretation of in

vitro experiments. Trivalent arsenic has a high affinity

for thiols, which makes glutathione an attractive ligand.

As part of my thesis project, I show for the first time that

AS(SG)3 can be isolated from a biological matrix and prove

that AS(SG)3 is one the species actually present in the

body.

GSH is fairly stable in aqueous solution, however GSH

is rapidly depleted in lung homogenates (Fig. 20) by

oxidation and binding to proteins. This decreases the

amount of GSH available and disrupts the labile arsenic-

glutathione complex. This explains why AS(SG)3

124

concentration decreases rapidly in lung homogenates but not

in aqueous solution. (Fig. 24). Unlike mercury, which forms

very strong bonds with GSH, As(SG)3 is a labile complex.

Small changes in equilibrium will disrupt the complex. As

excess GSH is oxidized, the complex is driven to dissociate.

When As (III) is used as a substrate for complex formation,

large amounts of complex are foirmed very quickly and slowly

dissipate. When As(V) is used as a substrate, 2 GSH are

required to reduce As(V) to As(III) before complexation can

occur. This reduces the rapidly dwindling supply of GSH and

accounts for the low formation of AS(SG)3 from As(V) (Fig.

24) .

Complex formation could significantly affect transport

of trivalent arsenic. As(III) is a neutral species at

physiological pH and appears to freely diffuse across

membranes (Ginsburg, 1965; Huang and Lee, 1996). As(SG)3 is

a charged species at physiological pH and will have to cross

membranes via a specific carrier mediated pathway. Several

studies have investigated the influence of GSH on As(III)

transport. Huang and Lee (1996) treated KB oral carcinoma

cells with mersalyl acid, a sulfhydryl modifier that does

not cross membranes, as well as NaN3 and KCN, which inhibit

active transport. These treatments significantly reduced

the uptake of As(V), but had no effect on As (III) uptake.

This indicates that As (III) uptake in this system is not

125

active and does not involve extracellular glutathione or

exofacial thiols. Hovever, these experiments were conducted

in RPMI media, which only contains Img/L, or 3.2|iM, reduced

glutathione, while arsenic concentrations were 200piM. Very

little of the arsenite would be complexed under these

conditions, therefore uptake of As(III) ion, not As(SG)3

complex was being measured. This is not likely to reflect

the in vivo situation, where nearly all of the arsenite will

be complexed. Georis et al. (1990) found that As(III)

uptake in liver slices was reduced by about 25% by

pretreatment with BSO to deplete intracellular thiols.

Uptake returned to normal with the addition of GSH to the

media, indicating a role for reduced thiols in As(III)

uptake. The conflicting results indicate the need for

further study on the role of extracellular thiols in As(III)

uptake. This is especially important for the uptake of

inhaled arsenic as lung epithelial lining fluid (ELF)

contains 430|iM GSH. This value is increased to 775|iM in

smokers (Cantin et al., 1987). It is likely that chronic

inhalation of arsenic would also increase ELF GSH content,

as previous work has shown that administration of arsenic

causes rebound increases in GSH (Rosner, 1989).

Although As(SG)3 has not been isolated from biological

systems, it has been implied to have multiple biological

effects. Gyurasics et al. (1991) reported that i.v.

126

injection of As(III) or As(V) increased the excretion of

NPSH in bile of rats. They attributed this to export of an

unstable As-GSH complex. As(86)3 has a mass of 994 and is

likely to excreted in the bile, as glutathione complexes

with a mass greater than 325 are largely excreted in bile

(Klaassen et al., 1981). Wang et al. (1996) found that

ethacrynic acid and Cibacron blue, inhibitors of glutahione-

S-transferases, decreased arsenic efflux from arsenic

resistant Chinese hamster cells. They determined that

As (III) was the form of arsenic exported, but found no

evidence for As(50)3. Given the lability of this complex,

it is likely that As(SG)3 was exported but decomposed to

As(III) and GSH during their chromatography. This work

implies a role for GSTs in formation of As(SG)3. While

AS(SG)3 forms readily from As(III) and GSH in solution, it

may require the activity of an enzyme to form it in

biological systems where the As(III) is likely to be bound.

Formation of As(SG)3 is an expensive process for the

lung. Potter and Tran (1993) determined that the half-life

of GSH in lung is 63.3 hours. The capacity of the lung to

synthesize GSH is 0.017 umol/g/hour. By comparison, the

half-life of GSH in the liver is 4.9 hours with a synthetic

capacity of 0.869 umol/g/hour. This means that GSH consumed

by complex formation is very hard to replace in the lung.

Glutathione reduces pentavalent arsenic [arsenate] to

trivalent arsenic [arsenite] in aqueous solution and

127

biological systems (Cullen et al., 1984; Scott et al., 1993;

Delnomdedieu et al, 1994; Winski and Carter, 1995).

Glutathione has been postulated to play an important role in

the methylation of inorganic arsenic as well as its

reduction (Buchet and Lauwerys, 1988; Thomas et al, 1995).

Trivalent arsenicals are methylated much more rapidly than

pentavalent arsenicals. However, much of the trivalent

arsenic is protein bound, which will compete with the

methylation reaction. Glutathione may act as a reductant to

reduce pentavalent arsenicals for more rapid reduction. It

may also act by forming a complex with trivalent arsenicals

that prevents them from binding to proteins and makes the

arsenic more available for methylation. As(SG)3 may

actually be the substrate for the methyltransferase, and

have to be formed prior to methylation. It is unclear from

our experiments if As(SG)3 is a better substrate for

arsenite methyltransferase than arsenite due to the presence

of GSH in the assay incubations. This poses some difficulty

because thiols are required for activity of the

methyltransferase.

4.2 Toxicity of arsenic species

It is known that the effect of arsenic in many tissues

is dependent on its chemical form. I hypothesized that the

128

same would be true in the lung. To investigate this, I

determined the effect of arsenite [As(III)], arsenate

[As(V)], MMA, DMA, and arsine (ASH3) on cell viability, heat

shock protein induction, and DNA single strand breaks.

These endpoints all have relevance to carcinogenesis, the

major toxic effect of arsenic in the lung.

The effect of arsenicals on BEAS-2B cell viability was

dependent on the form of arsenic. The rank order of

toxicity from the arsenicals was similar to that observed in

other systems (Klaassen, 1993). The LC50 for arsenite and

arsenate in BEAS-2B cells is similar to that observed for

these arsenicals in cell lines from other tissues.

Therefore, tracheal epithelial cells are not particularly

susceptible to toxicity from inorganic arsenic. This means

that these lung cells must be capable of detoxifying arsenic

and destroys the notion that the lung is particularly

susceptible to arsenic toxicity.

In hamster lung slices, toxicity was also dependent on

chemical form of the arsenic administered. Arsenate

produced very little toxicity while arsine and arsenite

produced significant toxicity. The toxicity observed by

potassium leakage was reflected in the damage observed

histologically. However, the amount of arsenic accumulated

was equal in arsenate and arsenite treated slices, despite

the differences in toxicity. This discrepancy must indicate

differences in the way that arsenic is affecting the cells

129

within the slice depending on the form of arsenic

administered. These differences may be explained by the

metabolism of each arsenical.

Pulmonary toxicity observed in human exposures to

arsine is restricted to edema and tachypnea (Hocken and

Bradshaw, 1970), however, arsine produced significant

toxicity in cultxured cells and lung slices. No long term

toxicity was observed in these patients. There are several

possible explanations for this discrepancy. Arsine rapidly

oxidizes to arsenite in lung homogenates. It is likely that

the arsenite produced from arsine is accumulating and

producing the toxicity observed. These are closed systems,

so arsenic will not be cleared as it is in vivo. In lung

slices, the agar used to maintain the lung structure may

have limited access of the arsenicals to the tissue. Arsine

is lipophillic and may diffuse more rapidly through the

agar, producing the higher concentrations of arsenic

observed in arsine treated lung slices compared to arsenite

and arsenate.

Arsenicals also cause effects on gene transcription.

One of the best characterized effects is on the induction of

heat shock, or stress response, proteins. Hsp32, also known

as heme oxygenase 1, is a stress response protein that

catalyzes the degradation of heme to biliverdin. It is

known to be strongly induced by heavy metals, including

arsenite, but not heat shock (Yoshida et al., 1988; Taketani

130

et al., 1989). It is clear that arsenic can induce this

protein, but due to metabolism of arsenic, it is unclear

what form of arsenic is active or the pathway involved in

arsenic induction of hsp32. Our results demonstrate that

only arsenite, arsenate, and arsine are capable of inducing

hsp32, but not MMA or DMA. Arsenate is reduced to arsenite

in the lung and probably causes hsp32 induction by

conversion to arsenite. Because arsine is oxidized to

arsenite in the lung, arsine may be inducing hsp32 by

oxidation to arsenite. However, arsine produces greater

induction of hsp32 than arsenite (Fig. 32) . This may

indicate that arsine induces hsp32 by a different mechanism.

Oxidant stress has also been shown to induce hsp32

(Applegate et al., 1991). Arsine has been hypothesized to

form radicals and may cause hsp32 induction by this pathway.

It is also possible that the lipophillicity of arsine allows

it to reach more critical areas before being oxidized to

arsenite. This may produce higher arsenite concentrations

at critical sites than arsenite treatment.

The fact that arsenite is active in induction of hsp32

leads to the idea that thiols are involved. Recent work by

Cavigelli et al. (1996) has shown that arsenite inhibits JNK

phosphatases, enzymes that have a critical cysteine residue.

Inhibition of JNK phosphatases causes AP-1 activation.

Hsp32 induction is mediated by AP-1 activation (Camhi et

131

al., 1995). Therefore it is likely that arsenite is

inducing hsp32 by activating AP-l.

This has significance because many tumor promoters lead

to AP-l activation. It is suggested that arsenic causes

tumor promotion by induction of AP-l, c-fos, and c-jun via

inhibition of JNK phosphatases (Cavigelli et al., 1996).

This would indicate that methylated arsenicals would not

function as tumor promoters by inducing AP-l because they do

not induce hsp32 expression.

While arsenicals are not mutagenic, they are

clastogenic. They can produce genetic alterations including

sister chromatid exchange and single strand breaks.

Previous work has shown that high concentrations of DMA

produces DNA single strand breaks in lung. Because DMA is a

metabolite of other arsenicals in mammals, I investigated

the ability of other arsenicals to produce single strand

breaks in BEAS-2B cells. Among the arsenicals tested, only

DMA was capable of inducing SSB. This agrees with the work

of Yamanaka et al. (1989) who showed that high doses of DMA

induce SSB in the lung. It is not clear why DMA produces

this damage, while other arsenicals do not.

Yamanaka has put forward evidence to suggest that DMA

causes damage by being converted to dimethylarsine and

further to the dimethylarsinoperoxyl radical (Yamanaka et

al., 1990). This leads to possible fojrmation of DNA adducts

of DMA, followed by incision repair that causes abasic sites

132

Which go on to form either single strand breaks or DNA-

protein crosslinks (Yamanaka et al., 1993; Yamanaka at al.,

1995).

Single strand DNA breaks may be caused by reduction of

DMA to dimethylarsine as proposed by Yamanaka et al. (1989).

However, reduction of DMA to dimethylarsine is a four

electron reduction requiring a potent reductant. The data

presented by Yamanaka et al. (1989) indicates that extremely

small amounts of dimethylarsine are produced from DMA. This

would limit dimethylarsinoperoxyl radical production to even

smaller amounts. The body is very efficient at detoxifying

radicals before they damage DNA, so a radical produced in

the minute quantities suggested by this data seems to have

little chance of producing the observed effects.

There are some differences between DMA and other

arsenicals. DMA has a pKa of 6.2. This means that

approximately 10% of this compound will be uncharged at

physiological pH, allowing diffusion across membranes. This

will be enhanced by the presence of the two methyl groups on

DMA. DMA is capable of being reduced to DMAs(III) by thiols

(Cullen et al., 1984). This is only a two electron

oxidation that occurs with thiol compounds found in the

body. The earlier comparison of methylation rates indicated

that MMA is more readily reduced than As(V) . If this is

true, DMA is likely to be even easier to reduce, due to the

presence of a second methyl group. Because DMA has two

133

methyl groups attached to the arsenic atom, it has only one

available binding site. In this respect DMAs(III) will

behave significantly differently than As(III). It will not

be tightly bound to vicinal dithiols and may undergo

oxidation/reduction cycling that could produce active oxygen

species responsible for the strand breaks.

Even still, other arsenicals are metabolized to DMA,

why don't they produce the same effects after metabolism?

The answer may be a balance of metabolism and toxicity.

High concentrations of DMA were required to produce single

strand breaks. If similar concentrations of arsenite are

added, so that enough DMA could be formed to cause this

effect. The cells will die of arsenite toxicity long before

strand breaks occur.

4.3 Correlation of metabolism and toxicity. Mann et

al. (1996) developed a physiologically based-pharmacokinetic

(PB-PK) model of arsenic metabolism. Because this model was

based on liver metabolism and whole body excretion data, it

cannot provide information on the species of arsenic

actually present in the lung after an inhalation exposure.

Our model is based solely on lung metabolism and provides a

profile of arsenic species present in the lung after

inhalation. Our model is not physiologically based and does

not account for clearance.

The toxicity studies clearly show that the effects of

arsenic in the lung are dependent on the chemical form of

134

arsenic. Therefore, it should be possible to correlate

observed effects with the concentration of the "active" form

of arsenic produced by metabolism. Using the information

collected from the metabolism studies, a computer model of

pulmonairy arsenic metabolism was constructed. The

calculated metabolite concentrations matched the observed

concentrations very well for arsenite and arsenate

metabolism, indicating that the model was accurate (Fig. 34

and 35).

The question that arose from this work was whether

arsine disposition could be explained solely by its

metabolism to arsenite. Arsenite is the first stable form

of arsenic to which arsine may be converted. The model of

pulmonary metabolism of arsine was applied assuming that

metabolism must first occur by conversion to arsenite. This

model produced discrepancies in the formation of MMA and

As(V) suggesting that arsine disposition cannot be explained

solely by oxidation to arsenite.

The formation of methylated metabolites from arsine

could proceed by two courses. Arsine could be oxidized to

arsenite and then methylated or it could be methylated

directly. The simulation was run assuming that only

arsenite could be methylated. The calculated MMA

concentrations were lower than the observed MMA

concentrations. This indicates that arsine may be directly

methylated. Arsine may lose a proton, forming AsH2~, which

135

can react with "*"0113 to form methylated derivatives. It

could also be an error in the simulation and proof must come

from isolation of methylated arsines.

Another interesting discrepancy was found in the

calculated As(V) production. The simulation only produced

As(V) by the oxidation of As (III) . The calculated As(V)

concentrations were much lower than observed, indicating

that arsine may oxidize directly to As(V).

The ultimate goal was to correlate observed effects

with specific arsenicals. To do this, toxicity data from

cells and metabolism data from homogenates must be combined.

There are several possible problems with this approach: 1)

uptake must not be a limiting factor and 2) metabolism in

BEAS-2B cells must be similar to that in homogenates. These

assximptions are made in the following calculations.

The model of pulmonary arsenic was used to approximate

the concentration of various forms of arsenic in BEAS-2B

cells treated with arsenicals. Not all arsenicals produced

effects, therefore, it was possible to assume an "active"

species of arsenic and correlate observed effects with

concentration of "active" arsenic species present after

various treatments.

In cell viability experiments, arsenite was more toxic

than arsenate or arsine. Because arsenite is produced from

arsenate and arsine, it is likely that the arsenite produced

during the course of the experiment accounted for the

136

toxicity produced by arsenate and arsine. To test this

hypothesis, the concentration of arsenite present 24 hours

after treatment with 300nM As(V) or ImM arsine was

calculated. In cells treated with 300(iM As(V) for 24 hours,

it was calculated that the arsenite concentration would be

around 75|iM. Toxicity observed from 75|iM arsenite is very

similar to that observed with 300nM As(V), suggesting that

the reduction of As(V) to As(III) accounts for the toxicity

produced by As(V). According to the model, cells treated

with ImM arsine for 24 hours would have an arsenite

concentration of around 185(iM. Treatment with this

concentration of arsenite would cause somewhat more toxicity

than is produced by ImM arsine. In this case, oxidation to

arsenite overestimates toxicity from arsine. This

discrepancy may be explained several ways. The model of

arsine metabolism may overestimate the production of

arsenite from arsine in cells due to errors in the model.

The other possibility is that cells can detoxify arsenite

produced by oxidation from arsine more efficiently than

arsenite produced by reduction of arsenate. This could

happen if arsine and arsenate are metabolized in different

compartments within the cell and these compartments handle

arsenite differently.

137

Arsenite, arsenate and arsine caused induction of

hsp32. Again, it is likely that arsenite is the form of

arsenic producing this effect. The model was used to

calculate the amount of arsenite produced during 4 hour

treatment with arsenicals. As seen in table 4a, induction

of hsp32 by arsenate correlates well with formation of

arsenite. However, induction of hsp32 by arsine does not

correlate well using this model. As discussed earlier,

arsine and arsenite may induce hsp32 by different mechanisms

or arsine may have a different distribution within the cell

that allows it to more efficiently induce hsp32.

BEAS-2B cells treated with ImM DMA for 12 hours had

significantly increased DNA single strand breaks. However,

treatment of cells with ImM MMA produced no effects. The

DMA concentration attained during a 12 hour incubation with

ImM MMA is only 25|iM. This concentration of DMA would

produce no effects on DNA single strand breaks.

Two interesting points arise from these correlations.

If the ability of every form of arsenic to produce an effect

is tested, it is possible to correlate effects to the

concentration of the "active" form of arsenic produced by

metabolism. This correlation is effective for all forms of

arsenic except arsine. This suggests that the metabolism

and disposition of arsine is different than other

arsenicals.

138

SUMMARY AND CONCLUSIONS

Reduction, oxidation, methylation, and complexation of

arsenic occurs in lung homogenates. Reduction in lung

homogenates is not strictly chemical, as homogenates reduced

arsenic faster than GSH alone. Supplementing homogenates

with physiological concentrations of GSH further increased

the rate of reduction, indicating the importance of reduced

thiols in the reduction process. Oxidation also occurred in

lung homogenates. The extent of arsenite oxidation

increased at higher concentrations, perhaps overwhelming

competing reductive processes. Arsine oxidizes rapidly in

lung homogenates to arsenite and arsenate.

Arsenic was methylated to mono- and dimethylated forms

of arsenic by lung cytosol. The rate of arsenic methylation

in lung was similar to that observed in liver, indicating

that the lung is not uniquely susceptible to arsenic

toxicity due to deficiencies in methylation. Arsine formed

methylated arsenic derivatives in lung cytosol preparations.

Like As(III), methylation of arsine is enzymatic as no

methylation occurred without protein.

Arsenite-glutathione complexes were isolated from lung

homogenates treated with arsenite and arsenate. The effect

of complex formation on metabolism and toxicity of arsenite

is unclear. However, complex formation may have a

139

significant impact on arsenite transport and should be

considered in future work.

The effects of arsenicals on lung cells was examined

and found to be dependent on the chemical form of arsenic.

Determining the ability of various forms of arsenic to

produce an effect allowed determination of an "active" form

of arsenic.

Using the data collected in the metabolism studies, a

model of arsenic metabolism in the lung was constructed.

This model was used to correlate the effects of various

arsenicals with metabolism to "active" forms. The

correlation was effective in explaining the effects of

arsenicals except for arsine. The model showed that the

metabolism and disposition of arsine is not explained solely

by oxidation to arsenite and subsequent oxidation or

methylation. Arsine appears to be oxidized directly to

As(V) and also be methylated directly to MMA.

APPENDIX A-Growing BEAS-2B cells

Media Components for addition to LHC-9 media from Bio-Fluids

human recombinant Epidermal Growth Factor Intergen 800-431-4505 Purchase, NY 10577

cat. no. 4110-80 lOO^g hrEGF

Bio-Fluids Inc. 800-972-5200 Rockville, MD

LHC-9 basal media List# 118 500ml $14.50 HBS (20mM HEPES buffered saline) List #340 100ml $4.75 P/E (ethanolamine/phosphoethanolamine) List # 353 20 ml $20.00 RA. (retinoic acid) List # 348 1ml $18.00 T3 (triiodothyronine) List # 354 1ml $19.00 hydrocortisone List # 346 20 ml $28.00 bovine pituitary extract (BPE) List #210 5ml $26.00 (order

3 or 4 at a time)

Sigma Chemical Co.

Insulin 12767 lOOmg $81.00 transferrin T-0519 lOOmg $41.30

human recombinant EGF E9644 lOOjig $51.45 epinephrine E4250 Ig $9.80 gentamycin G1397 10ml $36.75

Making LHC-9 media from LHC basal media

Stock solutions:

BSA stock

100 mg BSA 100 ml HBS Store at 4°C

141

Stock 4

0.042g FeS04-7H20 12.2g MgCla'eHzO 0.441g CaCl^-2H20 O.SmL conc. HCl Bring to IL with ddH20. Filter sterilize and store at room temperature.

Stock 11

0.863g ZnS04-7H20 Bring volume to IL with ddHaO. Filter sterilize and store at room temperature.

Calcium stock (116mM)

1.29g CaCla (anhydrous) Bring volume to lOOmL with ddHaO. Filter sterilize and store at room temperature.

Insulin stock (0.35mM)

30.0 mg insulin 15.0 mL 4mM HCl Store at 4°C

EGF stock (0.825|iM)

100 |4.g epidermal growth factor 2mL BSA stock 18.0 mL HBS store at -20°C

TF stock

500mg transferrin (human) 10 mL BSA stock 90 mL HBS filter sterilize and store at -20°C

Epinephrine stock

lOOmg epinephrine 100 mL lOmM HCl Store at -70°C

142

Retinoic acid stock (ImL)

IpiL retinoic acid (Biofluids-3. 3mM)

999^iL DMSO Wrap in aluminum foil, store at -70°C

Trace elements solution

ImL selenium solution (3.0mM) 52mg NaSeOs lOOmL ddHzO

ImL manganese solution (O.lmM) 1.26mg MnCl2-4H20 lOOmL ddHaO

ImL silicone solution 1.42mg NazSiOj-SHzO lOOmL ddHaO

ImL molybdenum solution 12.4mg (NHJ 6Mo7024 "4H20 lOOmL ddH20

ImL vanadium solution 5.9mg NH4VO3 lOOmL ddH20

ImL nickel solution 1. 3mg NiS04. 6H2O lOOmL ddH20

ImL tin solution 1. Img SnCl2.2H20 lOOmL ddH20

ImL concentrated HCl

Bring volume to IL with ddH20, filter sterilize and store at room temperature.

143

Add the following to 500mL LHC basal media to make LHC-9 media:

5mL Stock 4 0.5mL Stock 11 0.35mL Calcium stock 1.25mL Insulin stock O.lmL hydrocortisone (Biofluids-lOmM) O.SmL EGF stock l.OmL Transferrin stock 2.5 mL P/E Stock (Biofluids-0. ImM) 0.25mL Epinephrine stock O.OSmL Retinoic acid stock O.OSmL T3 (Biofluids-1.OmM) 5.OmL Trace elements solution O.SmL 50mg/ml gentcimycin 2.5mL Bovine pituitary extract 2.5mL lOOOOU/ml Pen/Strep

Filter sterilize and store at 4°C. Remove only what is needed from bottle, as repeated warming and cooling will degrade some media components.

Coating flasks

Cells must be plated on plastic vessels coated with fibronectin/vitrogen/BSA solution

The coating solution is made by mixing

1ml Vitrogen-100 (Celtrix, Santa Clara, CA) 10ml lOx BSA (lOx = Img/ml) Sigma Albiamin, Bovine fraction

V 5g $24.45 1 vial fibronectin (1 mg in 1ml in vial) Cal-Biochem

Bovine Plasma Fibronectin Cat no. 341631 Img $45.00 lOOml LHC-9 basal media

mix and filter before use

Coating flasks is done by adding minimal volume of coating solution (lml/75cm^), sloshing around plate to cover all areas, and allowing to sit for 10-15 minutes before removing liquid.

144

Plating and growing cells

Cells must be plated at a high enough density to provide cell-cell interactions. Cells will die if they remain confluent for more than 24 hours Cells should be split about 1:4 for subculturing Cells that were confluent and are split 1:4 should be confluent again in about 3 days

Subculturing cells

Remove media, rinse with PBS-PD Cells are removed from plates by Trypsin/EDTA (0.5% trypsin) +1% PVP solution Add 30% of T/E vol of SBTI (soybean trypsin inhibitor Img/ml in PBS-PD filtered before use)

-for example add 300(il SBTI to 1ml of T/E to stop trypsin action Spin cells down (5 minutes at 2000RPM) Resuspend in PBS-PD Spin down again Resuspend in modified LHC-9 media and plate

PVP is from Bio-Fluids List # 345 20 ml $7.00 SBTI is from Boehringer-Mannheim Cat. no. 109 886 50mg $31.00

PBS-PD is Calcium/Magnesium free PBS to make IL dissolve:

8g NaCl 0.2g Kcl 0.2g KH2PO4 1.15g NazHPO^

in water, pH to 7.4, bring to IL, and sterilize by

autoclaving. Store at 4°C.

145

REFERENCES

American Conference of Governmental Industrial Hygienists. (1982). TLVs; Threshold Limit Values for Chemical Substances in Work Air Adopted by ACGIH for 1982.

Aposhian. H.V. 1991. Biochemical toxicology of arsenic. Reviews of Biochemical Toxicology.

Applegate, L., Luscher, A., and Tyrrell, R. 1991. Induction of heme oxygenase: a general response to oxidant stress in cultured mammalian cells. Cancer Res. 51:974-978.

Axelson, O., Dcihlgren, E., Jansson, C.D., and Rehnlund, S.O. 1978. Arsenic exposure and mortality: A case-referent study from a Swedish copper smelter. Br. J. Ind. Med. 35:8-15.

Berenblum, I. and Shubik, P. 1947. A new quantitative approach to the study of stages of carcinogenesis in the mouse's skin. Br. J. Cancer 1:384.

Bogdan, G., Sampayo-Reyes, A., and Aposhian, H. 1994. Arsenic binding proteins of mammalian systems: I. Isolation of three arsenite binding proteins of rabbit liver. Toxicology 93:175-193.

Boutwell, R.K. 1964. Some biological aspects of skin carcinogenesis. Proge. Exp. Tumor Res. 4:207.

Brown, J. and Kitchin, K. 1996. Arsenite, but not cadmium, induces ornithine decarboxylase and heme oxygenase activity in rat liver: relevance to arsenic carcinogenesis. Cancer Letters 98:227-231.

Buchet, J.P. and Lauwerys, R. 1988. Role of thiols in the in vitro methylation of inorganic arsenic by rat liver cytosol. Biochem. Pharmacol. 37(16):3149-3153.

Buetler, E. (1984) . Red Cell Metabolism: A Manual of Biochemical Methods 3rd Ed. Grune and Stratton, Inc. Orlando, FL.

Burnette, W.N. 1981. "Western Blotting": electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 112:195.

Challenger, F. 1945. Biological Methylation. Chem. Rev. 36:315-361.

146

Camhi, S., Alam, J., Otterbein, L., Sylvester, S., and Choi, A.M.K. 1995. Induction of heme oxygenase-1 gene expression by lipopolysaccharide is mediated by AP-1 activation. Am. J. Respir. Cell Mol. Biol. 13:387-398.

Cantin, A., North, S., Hubbard, R., and Crystal, R. 1987. Normal alveolar epithelial lining fluid contains high levels of glutathione. J. Appl. Physiol. 63(1):152-157.

Cavigelli, M., Li, W., Lin, A., Su, B., Yoshioka, K., and Karin, M. 1996. The tumor promoter arsenite stimulates AP-1 activity by inhibiting a JNK phosphatase. EMBO J. 15(22):6269-6279.

Choi, S., Chase, T. Jr., and Bartha, R. 1994. Enzymatic catalysis of mercury methylation by desulfovibrio desulfuricans LS. Appl. Environ. Micro. 60(4):1342-1346.

Cullen, W., McBride, B., and Reglinski, J. 1984. The reaction of methylarsenicals with thiols: some biological implications. J. Inorg. Chem. 21:179-194.

Davidson, C.I., Goold, W.D., Mathison, T.P, et al. 1985. Airborne trace elements in Great Smoky Mountains, Olympic, and Glacier National Parks. Environ. Sci. Technol. 19:27-35.

Delnomdedieu, M., Basti, M., Otvos, J., and Thomas, D. 1994b. Reduction and binding of arsenate and dimethylarsinate by glutathione: a magnetic resonance study. Chem.-Biol. Interactions 90:139-155.

Delnomdedieu, M., Basti, M., Styblo, M., Otvos, J., and Thomas, D. 1994a. Complexation of arsenic species in rabbit erythrocytes. Chem. Res. Toxicol. 7:621-627.

Enterline, P.E., Henderson, V.L., and Marsh, G.M. 1987. Exposure to arsenic and respiratory cancer: a reanalysis. Am. J. Epidemiol. 125:929-938.

Fisher, R., Shaughnessy, R., Jenkins, P., Austin, M., Roth, G., Gandolfi, A.J., and Brendel, K. (1995). Dynamic organ culture is superior to multiwell plate culture for maintaining precision-cut tissue slices: optimization of tissue slice culture, Part 1. Toxicology Methods 5(2), 99-113.

Foulds, L. 1965. Multiple etiologic factors in neoplatic development. Cancer Res. 25:1339-1347.

Fowler, B. and Weissberg, J. 1974. Arsine Poisoning. The New England Journal of Medicine 291(22):1171-1174.

147

Gilman, A., Goodman, L., Rail, T., and Murad, F., eds. 1985. Goodman and Gilman's The Pharmacological Basis of Therapeutics. 7^^ edition. Macmillan Publishing Co., New York. pp. 1614-1617.

Ginsbxirg, J. 1965. Renal mechanism for excretion and transformation of arsenic in the dog. Am. J. Physiol. 208(5):832-840.

Gresser, M.J. 1981. ADP-Arsenate: Formation by submitochondrial particles under phosphorylating conditions. J. Biol. Chen. 256:5981-5983.

Gubits, R.M. 1988. C-Fos mRNA levels are increased by the cellular stressors, heat shock, and sodium arsenite. Oncogene 3: 163-168.

Gyurasics, A., Varga, F., and Gregus, Z. 1991. Effect of arsenicals on biliary excretion of endogenous glutathione and xenobiotics with glutathione dependent hepatobiliary transport. Biochem. Pharmacol. 41(6/7):937-944.

Healy, S.M., Zakharyan, R.A., and Aposhian, H.V. 1997. Enzymatic methylation of arsenic compounds: IV. In vitro and in vivo deficiency of the methylation of arsenite and monomethylarsonic acid in guinea pigs. Mut. Res. 386:229-239.

Hocken, A.G. and Bradshaw, G. (1970). Arsine Poisoning. Brit. J. Indust. Med. 27:56-60.

Hopenhayn-Rich, C., Biggs, M., Kalman, D., Moore, L., and Smith, A. 1996. Arsenic methylation patterns before and after changing from high to lower concentrations of arsenic in drinking water. Environ. Health Perspec. 104(11):1200-1207.

Huang, R-N. and Lee, T-C. 1996. Cellular uptake of trivalent arsenite and pentavalent arsenate in KB cells cultured in phosphate-free medium. Toxicol. Appl. Pharmacol. 136:243-249.

Jacobson-Kram, D. and Montalbano, D. 1985. The reproductive effects assessments•s group report on the mutagenicity of inorganic arsenic. Environ. Mutagen. 7:767-804.

Jarup, L. and Pershagen, G. 1991. Arsenic esposure, smoking, and lung cancer in smelter workers- A case-control study. Am. J. Epidem. 134(6):545-551.

148

Jarup, L., Pershagen, G., and Wall, S. 1989. Cumulative arsenic exposure and lung cancer in smelter workers: a dose-response study. Am. J. Jnd. Med. 15:31-41.

Ji, G. and Silver, S. 1992. Reduction of arsenate to arsenite by the ArsC protein of the arsenic resistance operon of Staphylococcus aureus plasmid pI258. Proc. Natl. Acad. Sci. 89:9474-9478.

Josephson, C., Pinto, S., and Petronella, S. 1951. Arsine: electrocardiographic changes produced in human poisoning. Arch. Indust. Hyg. 4:43-52.

Klaassen, C.D., ed. 1996. Casaret & Doull's Toxicology: The Basic Science of Poisons, ed. McGraw-Hill Co., Inc. pp. 696-697.

Klaassen, C.D., Eaton, D.L., and Cagen, S.Z. 1981. Hepatobiliary disposition of xenobiotics. In Bridges, J.W. and Chasseaud, L.F. (eds.): Progress in Drug Metabolism. John Wiley and Sons, Inc. New York. p.1-75.

Krause, T., Einhaus, M., Holz, O., Meissner, R., Baumgartner, E., and Rudiger, H. 1993. A novel technique for the detection of DNA single-strand breaks in human white blood cells and its combination with the unscheduled DNA synthesis assay. Intl. Arch. Occup. Environ. Health 65:77-82.

Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of a bacteriophage T4. Nature 227:680.

Lau BP. Michalik P. Porter CJ. Krolik S. Identification and confirmation of arsenobetaine and arsenocholine in fish, lobster and shrimp by a combination of fast atom bombardment and tandem mass spectrometry. Biomedical & Environmental Mass Spectrometry. 14 (12) :723-32, 1987 Dec.

Lechner, J.F. and LaVeck, M.A. 1985. A serum-free method for culturing normal human bronchial epithelial cells at clonal density. J. Tissue Culture Methods 9(2):43-48.

Lee, A.M. and Fraumeni, J.F. 1969. Arsenic and respiratory cancer in man: An occupational study. J. Natl. Cancer Inst. 42:1045-1052.

Lee, T.C., Oshimura, M., and Barrett, J.C. 1985. Comparison of arsenic-indcued cell transformation, cytotoxicity, mutation, and cytogenetic effects in Syrian hamster cells in culture. Carcinogenesis 6:1421-1426.

149

Lee, T.C., Tanaka, N., Lamb, P.W., Gilmer, T.M., and Barrett, J.C. 1988. Induction of gene amplification by arsenic. Science 241:79-81.

Lee-Feldstein, A. 1986. A ciomulative exposure to arsenic and its relationship to respiratory cancer among copper smelter employees. J. Occup, Med. 28:296-302.

Leman, S., Clarkson, T., and Gerson, R. 1983. Arsenic uptake and metabolism by liver cells is dependent on arsenic oxidation state. Chem.-Biol. Interactions 45:401-406.

Li, J-H. and Rossman, T. 1989. Inhibition of DNA ligase activity by arsenite: a possible mechanism of its co-mutagenesis. Molec. Toxicol. 2:1-9.

Li, J-H. and Rossman, T. 1991. Comutagenesis of sodium arsenite with ultraviolet radiation in Chinese hamster V79 cells. Biol. Metals 4:197-200.

Lowe, T., Chen, Q., Fernando, Q., Keith, R., and Gandolfi, A.J. (1993). Elemental analysis of renal slices by proton induced x-ray emission. Environmental Health Perspectives 101, 302-308.

Mann, S., Droz, P., and Vahter, M. 1996. A physiologically based pharmacokinetic model for arsenic exposure II. Validation and application in humans. Toxicol. Appl. Pharmacol. 140:471-486.

Marafante, E., Vahter, M., and Envall, J. 1985. The role of methylation in the detoxication of arsenate in the rabbit. Chem.-Biol. Interactions. 56:225-238.

Montgomery, H. 1935. Arsenic as etiologic agent in certain types of epithelioma. Arch. Dermat. Syph. 32:218.

Moore, L.E., Smith, A.H., Hopenhayn-Rich C., Biggs, M.L., Kalman, D.A., and Smith M.T. 1997. Micronuclei in exfoliated bladder cells among individuals chronically exposed to arsenic in drinking water. Cancer Epidemiology, Biomarkers & Prevention. 6(1):31-6.

Muehrcke, R., and Pirani, C. 1968. Arsine-induced anuria: a correlative clinicopathological study with electron microscopic observations. Ann. Int. Med. 68:853-866.

Nakano, M., Yamanaka, K., Hasegawa, A., Sawamura, R., and Okada, S. 1992. Preferential increase of heterochromatin in venular endothelium of lung after administration of dimehtylarsinic acid, a major metabolite of inorganic arsenics. Carcinogenesis 13(3) :391-393.

150

Neubauer, O. 1947. Arsenical cancer. Brit. J. Cancer 1:192.

Newman, J.A., Archer, V.E., Saccomanno, G., et al., Occupational carcinogenesis. Histologic types of bronchogenic carcinoma among members of copper mining and smelting communities. Ann. N.Y. Acad. Sci. 271:260-268.

Nunoshiba, T. and Nishioka, H. 1987. Sodium arsenite inhibits spontaneous and induced mutations in Escherichia Coli. Mutat. Res. 184:99-105.

Osborne, F.H. and Ehrlich, H.L. 1976. Oxidation of arsenite by a soil isolate of Alcaligenes. J. Appl. Bacterial. 41:295-305.

Ott, M.G., Holder, B.B., and Gordon, H.L. 1974. Respiratory cancer and occupational exposures to arsenicals. Arch. Environ. Health 29:250-255.

Oya-Ohta, Y., Kaise, T., Ochi T., Induction of chromosomal aberrations in cultured human fibroblasts by inorganic and organic arsenic compounds and the different roles of glutathione in such induction. Mutation Research. 357(1-2);123-9, 1996 Oct 25.

Paris, J.A. 1820. Pharmacologia III. W. Phillips, London, pp. 132-134; cited by Buchanan, W.D. Toxicity of arsenic compounds. Elsevier, New York. 1962.

Pernis, B., and Magistretti, M. 1960. A study of the mechanism of acute hemolytic anemia from arsine. Med. Lavoro 51:37-41.

Pershagen, G., Nordberg, G., and Bjorkland, N-E. 1984a. Carcinomas of the respiratory tract in hamsters given arsenic trioxide and/or benzo(a) pyrene by the pulmonary route. Environ. Res. 34:227-241.

Pershagen, G., Bergman, F., Klominek, J., Damber, L., and Wall, S. 1987. Histological types of lung cancer among smelter workers exposed to arsenic. Br. J. Ind. Med. 44:454-458.

Pitot, H.C. and Dragan, Y.P. 1994. Chemical induction of hepatic neoplasia. In Arias, I., Boyer, J., and Fausto, N. (eds.): The liver: biology and pathobiology, 3 ed. New York: Raven Press, pp.14 67-1495.

Potter, D. and Tran, T-B. 1993. Apparent rates of glutathione turnover in rat tissues. Toxicol. Appl. Pharmacol. 120:186-192.

151

Rabano, E.S., Castillo, N.T., Torre, K.J., and Solomon, P.A. 1989. Speciation of arsenic in ambient aerosols collected in Los Angeles. JAPCA 39(l):76-80.

Roehm, N., Rodgers, G., Hatfield, S., Glasebrook, A. 1991. An improved colorimetric assay for cell proliferation and viability utilizing the tetrazolium salt XTT. J. Immunological Methods 142:257-265.

Rossman, T.G. 1981. Effect of metals on mutagenesis and DNA repair. Environ. Health Perspect. 40:189-195.

Shirachi, D.Y., Tu, S.H., and McGowan, J.P. 1987. Carcinogenic effects of arsenic compounds in drinking water. Cincinnati, OH: US EPA, HERL. NTIS no. PB87-232542. EPA 600/1-87-007.

Scott, N., Hatlelid, K., MacKenzie, N., and Carter, D.E. 1993. Reactions of As(III) and As(V) species with glutathione. Chem. Res. Toxicol. 6(1):102-106.

Shalat, S., Walker, D., and Finnell, R. 1996. Role of arsenic as a reproductive toxin with particular attention to neural tube defects. J. Toxicol. Environ. Health 48:253-272.

Solomon, P.A. 1984. Collection and analysis techniques for studies of atmospheric arsenic species. Ph.D. Thesis. Dept. of Chemistry, Univ. of Arizona, Tucson, AZ.

Squibb, K.S. and Fowler, B.A. 1983. The toxicity of arsenic and its compounds, in Biological and Environmental Effects of Arsenic. B.A. Fowler, ed. Elsevier Press, Amsterdma, The Netherlands.

Stevenson, K., Hale, G., and Perham, R. 1978. Inhibition of pyruvate dehydrogenase multienzyme complex from Escerichia Coli with mono- and bifunctional arsenoxides. Biochemistry 17:2189-2192.

Thompson, D.J. 1993. A chemical hypothesis for arsenic methylation. Chem.-Biol. Interactions 88:89-114.

Tinwell, H., Stephens, S.C., and Ashby, J. 1991. Arsenite as the probable active species in the human carcinogenicity of arsenic: mouse micronucleus assays on Na and K arsenite, orpiment, and Fowler's solution. Environ. Health Perspec. 95:205-210.

Vahter, M. and Marafante, E. 1989. Intracellular ditribution and chemical forms of arsenic in rabbits exposed to arsenate. Biol. Trace Element Res. 21:233-239.

152

Van Wijk, R., Welters, M., Souren, J., Ovelgonne, H., and Wiegant, F. 1993. Senom-stimulated cell cycle progression and stress protein synthesis in C3H10T1/2 fibroblasts treated with sodiiim arsenite. J. Cell. Physiol. 155:265-272.

Voegtlin, C., Dyer, H., and Leonard, C. 1924. On the mechanism of the action of arsenic upon protoplasm. U.S. Public Health Rep. 38:1882-1912.

Wang, Z., Dey, S., Rosen, B., and Rossman, T. 1996. Efflux-mediated resistance to arsenicals in arsenic-resistant and -hypersensitive Chinese hamster cells. Toxicol. Appl. Pharmacol. 137:112-119.

Wanibuchi, H., Yamamoto, S., Chen, H., Yoshida, K., Endo, G., Hori, T., and Fvikushima, S. 1996. Promoting effects of dimethylarsinic acid on N-butyl-N-(4-hydroxybutyl) nitrosamine-induced urinary bladder carcinogenesis in rats. Carcinogenesis 17(11):2435-2439.

Weast, R.C., ed. (1974). CRC Handbook of Chemistry and Physics. 55th ed. Cleveland, Ohio: CRC Press, Inc. pp.B-69.

Weast, R.C., ed. (1976). CRC Handbook of Chemistry and Physics. 55th ed. Cleveland, OH: CRC Press, Inc. pp.B-9.

Webb, J.L. 1966. Enzvme and Metabolic Inhibtors. Vol. Ill, Academic Press, New York, pp. 595-793.

Welch, K. , Higgins, I., Oh, M., and Burchfiel, C. 1982. Arsenic exposure, smoking, and respiratory cancer in copper smelter workers. Arch. Environ. Health 37:325-335.

Wicks, M. Archer, V., Auerbach, O., and Kuschner, M. 1981. Arsenic exposure in a copper smelter as related to histological type of lung cancer. Am. J. Ind. Med. 2:25-31.

Winski, S. and Carter, D.E. 1995. Interactions of rat red blood cell sulfhydryls with arsenate and arsenite. J. Toxicol. Environ. Health 46:379-397.

Yamanaka, K., Hasegawa, A., Sawamura, R., and Okada, S. 1989. Dimethylated arsenic induce DNA strand breaks in lung via the production of active oxygen in mice. Biochem. and Biophys. Res. Comm. 165(1):43-50.

153

Yamanaka, K., Hasegawa, A., Sawamura, R., and Okada, S. 1991. Cellular response to oxidative damage in lun ginduced by the administration of dimethylarsinic acid, a major metabolite of inorganic arsenics, in mice. Toxicol. Appl. Pharmacol. 108:2 05-213.

Yamanaka, K., Hayashi, H., Kato, K., Hasegawa, A., and Okada, S. 1995. Involvement of preferential formation of apxirinic/apyrimidinic sites in dimethylarsinic-induced DNA strand breeJcs and DNA-protein crosslinks in cultxired alveolar epithelial cells. Biochem. and Biophys. Res. Comm. 207(1):244-249.

YamaneOca, K., Hoshino, M., Okamoto, M., Hasegawa, A., Sawamura, R., and Okada, S. 1990. Induction of DNA damage by dimethylarsine, a metabolite of inorganic arsenics, is for the major part likely due to its peroxyl radical. Biochem. and Biophys. Res. Comm. 168(1): 58-64.

Yamanaka, K., Ohtsubo, K., Hasegawa, A., Hayashi, H., Ohji, H., Kanisawa, M., and Okada, S. 1996. Exposure to dimethylarsinic acid, a main metabolite of inorganic arsenics, strongly promotes tumorigenesis intiated by 4-nitroquinoline l-oxide in the lungs of mice. Carcinogenesis 17(4):767-770.

Yamanaka, K., Tezuka, M., Kato, K., Hasegawa, A., and Okada, S. 1993. Crosslink formation between DNA and nuclear proteins by in vivo and in vitro exposure of cells to dimethylarsinic acid. Biochem. and Biophys. Res. Comm. 191(3):1184-1191.

Zakharyan, R., Wu, Y., Bogdan, G., and Aposhian, H. 1995. Enzymatic methylation of arsenic compounds: assay, partial purification, and properties of arsenite methyltransferase and monomethylarsonic acid methyltransferase of rabbit liver. Chem. Res. Toxicol. 8:1029-1038.